<html><body><div style="font-family: garamond,new york,times,serif; font-size: 12pt; color: #000000"><div>Please notice time change.<br></div><hr id="zwchr" data-marker="__DIVIDER__"><div data-marker="__HEADERS__"><br></div><div data-marker="__QUOTED_TEXT__"><div style="font-family: garamond,new york,times,serif; font-size: 12pt; color: #000000;" data-mce-style="font-family: garamond,new york,times,serif; font-size: 12pt; color: #000000;"><div><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;">Alexander Beatson will present his general exam on <span class="Object" id="OBJ_PREFIX_DWT6258_com_zimbra_date" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;">Wednesday, May 10</span>, 2017 at<strong> 4pm</strong> in CS 301. The members of his committee are: Han Liu (ORFE, adviser), Elad Hazan, and Arvind Narayanan.</div><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><br></div><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;">Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below.</div></div></div><div style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><br></div></div><div><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Title: Deep learning in some non-iid settings</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">In real-world settings observed data may often be non-iid. There might be differences between the training and deployment distributions, which we can model in the worst case as perturbation by an adversary [1], or the training data may be drawn from different distributions at different times or at different locations, and we might wish to learn from the global distribution without storing or communicating the data [12]. I will discuss some work on attacks on machine learners and on non-iid distributed deep learning with communication constraints.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;">Attacks on machine learners:</span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;"> One threat model for machine learning involves an adversary perturbing the training set. We show that even a “blind” adversary with no knowledge of the training data can harm learning (meaningfully lower-bound the learner’s effective sample size) via simple data injection strategies. Another threat model involves an adversary perturbing the examples observed in deployment. Linear and deep models are particularly vulnerable to such adversarial examples [2-4]. We show that deep generative models can be used to correctly classify adversarial examples or detect them as being far from the manifold of real data.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;">Non-iid distributed deep learning:</span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;"> “Federated learning” [12] features communication constraints and data distributed non-iid across nodes. We propose an asynchronous algorithm for this setting which uses the Hessian of the likelihood at each node to weight model averaging and compute proximal constraints. This may be derived as an iterated natural proximal method or an application of the Laplace approximation [9]. Experiments show this converges faster than other asynchronous algorithms when data are non-iid across nodes. We may approximate the algorithm by sending only a small fraction of the variable updates from each node, sampled according to diagonal of that node’s Hessian. This maintains improved convergence in the non-iid setting, and in both iid and non-iid settings reduces the communication volume needed to reach a given accuracy.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Reading List:</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;">Attacks on machine learners:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Overview of the field:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[1] </span><span class="Object" id="OBJ_PREFIX_DWT6236_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1606.06565" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Concrete problems in AI safety</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman, Dan Mané.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Adversarial examples in deep learning:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[2] </span><span class="Object" id="OBJ_PREFIX_DWT6237_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1412.6572" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Explaining and harnessing adversarial examples</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Ian J. Goodfellow, Jonathon Shlens, Christian Szegedy.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[3] </span><span class="Object" id="OBJ_PREFIX_DWT6238_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1602.02697" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Practical black-box attacks against deep learning systems using adversarial examples</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Nicolas Papernot, Patrick McDaniel, Ian Goodfellow, Somesh Jha, Z. Berkay Celik, Ananthram Swami.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[4] </span><span class="Object" id="OBJ_PREFIX_DWT6239_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1608.04644" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Towards Evaluating the Robustness of Neural Networks</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Nicholas Carlini, David Wagner.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Previously proposed defenses:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[5] </span><span class="Object" id="OBJ_PREFIX_DWT6240_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1703.09202" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Biologically inspired protection of deep networks from adversarial attacks</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Aran Nayebi, Surya Ganguli.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[6] </span><span class="Object" id="OBJ_PREFIX_DWT6241_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1704.08847" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Parseval Networks: Improving Robustness to Adversarial Examples</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Cisse Moustapha, Bojanowski Piotr, Grave Edouard, Dauphin Yann, Usunier Nicolas.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Deep generative models:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[7] </span><span class="Object" id="OBJ_PREFIX_DWT6242_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1406.2661" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Generative Adversarial Networks</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[8] </span><span class="Object" id="OBJ_PREFIX_DWT6243_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1312.6114" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Auto-Encoding Variational Bayes</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Diederik P Kingma, Max Welling.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;">Non-iid and distributed deep learning:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">The Laplace approximation: </span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[9] </span><span class="Object" id="OBJ_PREFIX_DWT6244_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="http://www.mitpressjournals.org/doi/pdf/10.1162/neco.1992.4.3.448" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">A practical Bayesian framework for backpropogation networks</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. David J. C. MacKay.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Using the Laplace approximation to estimating weight importances in deep learning:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[10] </span><span class="Object" id="OBJ_PREFIX_DWT6245_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="http://yann.lecun.com/exdb/publis/pdf/lecun-90b.pdf" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Optimal Brain Damage</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Yann LeCun, John S. Denker, Sara A. Solla.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[11] </span><span class="Object" id="OBJ_PREFIX_DWT6246_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1612.00796" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Overcoming catastrophic forgetting in neural networks</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. James Kirkpatrick, Razvan Pascanu, Neil Rabinowitz, Joel Veness, Guillaume Desjardins, Andrei A. Rusu, Kieran Milan, John Quan, Tiago Ramalho, Agnieszka Grabska-Barwinska, Demis Hassabis, Claudia Clopath, Dharshan Kumaran, Raia Hadsell.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-style: italic; vertical-align: baseline; white-space: pre-wrap;">Federated learning:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[12] </span><span class="Object" id="OBJ_PREFIX_DWT6247_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1602.05629" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Communication-Efficient Learning of Deep Networks from Decentralized Data</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. H. Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, Blaise Agüera y Arcas.</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[13] </span><span class="Object" id="OBJ_PREFIX_DWT6248_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="https://arxiv.org/abs/1610.05492" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Federated Learning: Strategies for Improving Communication Efficiency</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Jakub Konečný, H. Brendan McMahan, Felix X. Yu, Peter Richtárik, Ananda Theertha Suresh, Dave Bacon.</span></p><br style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial;"><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; font-weight: bold; vertical-align: baseline; white-space: pre-wrap;">Textbook:</span></p><p dir="ltr" style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;" data-mce-style="color: #000000; font-family: garamond, 'new york', times, serif; font-size: 16px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: #ffffff; text-decoration-style: initial; text-decoration-color: initial; line-height: 1.38; margin: 0px;"><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">[14] </span><span class="Object" id="OBJ_PREFIX_DWT6249_com_zimbra_url" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><a href="http://www.deeplearningbook.org/" target="_blank" style="color: #005a95; text-decoration: none; cursor: pointer;" data-mce-style="color: #005a95; text-decoration: none; cursor: pointer;"><span style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; background-color: transparent; text-decoration-line: underline; vertical-align: baseline; white-space: pre-wrap;">Deep Learning</span></a></span><span style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;" data-mce-style="font-family: 'times new roman'; color: #000000; background-color: transparent; vertical-align: baseline; white-space: pre-wrap;">. Ian Goodfellow, Yoshua Bengio, Aaron Courville.</span></p></div></div><br>_______________________________________________<br>talks mailing list<br>talks@lists.cs.princeton.edu<br>To edit subscription settings or remove yourself, use this link:<br>https://lists.cs.princeton.edu/mailman/listinfo/talks<br></div></div></body></html>