In a recently published article on Fortune.com He begins by noting that data science and analytics has been in the "dancing bear stage" and that it's time for it to "demonstrate its ability to materially impact health and disease." Much of the article focuses on parsing an Andreesen Horowitz podcast interview with Vas Narasimhan, chief executive officer at Novartis, and his views on the (limited) successes and (undoubtedly expensive) failures with data science and analytics deployments in pharma. Early in the article, Shaywitz quotes a paper by Dr. Sachin Jain, former CMIO at Merck and now CEO of CareMore Health: "The conference circuit is now exploding with "AI and pharma" conferences… consultants excitedly discuss pharma's digital transformation (and convince each pharma brand they're distinctly behind), and exuberant stories about the power of data and AI resound almost daily across social media." This question of the progress with applying artificial intelligence in pharma is not solely a digital transformation issue.

Tesla has announced it is taking orders for the mass-market Model 3, which has a starting price of $35,000 (£26,400). The company said it would close stores and shift all its sales to the internet in order to cut costs for the electric car. Its CEO, Elon Musk, said the move was essential to Tesla's survival. "People want to buy online." Musk also backed off of earlier guidance that the company would be profitable in all future quarters.

Private investment in artificial intelligence (AI) start-ups is growing, mainly in the US and China, according to an OECD analysis of Crunchbase data. Start-ups operating in the US account for the majority of AI start-up equity investments worldwide, while China has quickly become the second player globally in terms of AI equity investments in start-ups. In 2017, Chinese start-ups attracted 36 per cent of global AI investment in 2017, a significant jump from just 3 per cent in 2015. Source: OECD estimates based on Crunchbase (April 2018), www.crunchbase.com

Private investment in artificial intelligence (AI) start-ups is growing, mainly in the US and China, according to an OECD analysis of Crunchbase data. Start-ups operating in the US account for the majority of AI start-up equity investments worldwide, while China has quickly become the second player globally in terms of AI equity investments in start-ups. In 2017, Chinese start-ups attracted 36 per cent of global AI investment in 2017, a significant jump from just 3 per cent in 2015. Source: OECD estimates based on Crunchbase (April 2018), www.crunchbase.com

E-Commerce (E-Com) search is an emerging important new application of information retrieval. Learning to Rank (LETOR) is a general effective strategy for optimizing search engines, and is thus also a key technology for E-Com search. While the use of LETOR for web search has been well studied, its use for E-Com search has not yet been well explored. In this paper, we discuss the practical challenges in applying learning to rank methods to E-Com search, including the challenges in feature representation, obtaining reliable relevance judgments, and optimally exploiting multiple user feedback signals such as click rates, add-to-cart ratios, order rates, and revenue. We study these new challenges using experiments on industry data sets and report several interesting findings that can provide guidance on how to optimally apply LETOR to E-Com search: First, popularity-based features defined solely on product items are very useful and LETOR methods were able to effectively optimize their combination with relevance-based features. Second, query attribute sparsity raises challenges for LETOR, and selecting features to reduce/avoid sparsity is beneficial. Third, while crowdsourcing is often useful for obtaining relevance judgments for Web search, it does not work as well for E-Com search due to difficulty in eliciting sufficiently fine grained relevance judgments. Finally, among the multiple feedback signals, the order rate is found to be the most robust training objective, followed by click rate, while add-to-cart ratio seems least robust, suggesting that an effective practical strategy may be to initially use click rates for training and gradually shift to using order rates as they become available.

Human perception is structured around objects which form the basis for our higher-level cognition and impressive systematic generalization abilities. Yet most work on representation learning focuses on feature learning without even considering multiple objects, or treats segmentation as an (often supervised) preprocessing step. Instead, we argue for the importance of learning to segment and represent objects jointly. We demonstrate that, starting from the simple assumption that a scene is composed of multiple entities, it is possible to learn to segment images into interpretable objects with disentangled representations. Our method learns -- without supervision -- to inpaint occluded parts, and extrapolates to scenes with more objects and to unseen objects with novel feature combinations. We also show that, due to the use of iterative variational inference, our system is able to learn multi-modal posteriors for ambiguous inputs and extends naturally to sequences.

We present a novel technique for tailoring Bayesian quadrature (BQ) to model selection. The state-of-the-art for comparing the evidence of multiple models relies on Monte Carlo methods, which converge slowly and are unreliable for computationally expensive models. Previous research has shown that BQ offers sample efficiency superior to Monte Carlo in computing the evidence of an individual model. However, applying BQ directly to model comparison may waste computation producing an overly-accurate estimate for the evidence of a clearly poor model. We propose an automated and efficient algorithm for computing the most-relevant quantity for model selection: the posterior probability of a model. Our technique maximizes the mutual information between this quantity and observations of the models' likelihoods, yielding efficient acquisition of samples across disparate model spaces when likelihood observations are limited. Our method produces more-accurate model posterior estimates using fewer model likelihood evaluations than standard Bayesian quadrature and Monte Carlo estimators, as we demonstrate on synthetic and real-world examples.

Sleep arousals transition the depth of sleep to a more superficial stage. The occurrence of such events is often considered as a protective mechanism to alert the body of harmful stimuli. Thus, accurate sleep arousal detection can lead to an enhanced understanding of the underlying causes and influencing the assessment of sleep quality. Previous studies and guidelines have suggested that sleep arousals are linked mainly to abrupt frequency shifts in EEG signals, but the proposed rules are shown to be insufficient for a comprehensive characterization of arousals. This study investigates the application of five recent convolutional neural networks (CNNs) for sleep arousal detection and performs comparative evaluations to determine the best model for this task. The investigated state-of-the-art CNN models have originally been designed for image or speech processing. A detailed set of evaluations is performed on the benchmark dataset provided by PhysioNet/Computing in Cardiology Challenge 2018, and the results show that the best 1D CNN model has achieved an average of 0.31 and 0.84 for the area under the precision-recall and area under the ROC curves, respectively.

Differential privacy is becoming a standard notion for performing privacy-preserving machine learning over sensitive data. It provides formal guarantees, in terms of the privacy budget, $\epsilon$, on how much information about individual training records is leaked by the model. While the privacy budget is directly correlated to the privacy leakage, the calibration of the privacy budget is not well understood. As a result, many existing works on privacy-preserving machine learning select large values of $\epsilon$ in order to get acceptable utility of the model, with little understanding of the concrete impact of such choices on meaningful privacy. Moreover, in scenarios where iterative learning procedures are used which require privacy guarantees for each iteration, relaxed definitions of differential privacy are often used which further tradeoff privacy for better utility. In this paper, we evaluate the impacts of these choices on privacy in experiments with logistic regression and neural network models. We quantify the privacy leakage in terms of advantage of the adversary performing inference attacks and by analyzing the number of members at risk for exposure. Our main findings are that current mechanisms for differential privacy for machine learning rarely offer acceptable utility-privacy tradeoffs: settings that provide limited accuracy loss provide little effective privacy, and settings that provide strong privacy result in useless models. Open source code is available at https://github.com/bargavj/EvaluatingDPML.

We consider PAC learning for identifying a good item from subset-wise samples in \pl\, probability models, with instance-dependent sample complexity performance. For the setting where subsets of a fixed size can be tested and top-ranked feedback is made available to the learner each time, we give the first $(\epsilon,\delta)$-PAC best item algorithm with an instance-dependent sample complexity bound. The algorithm relies on a wrapper that uses a weaker PAC algorithm with worst-case performance guarantees to adapt to the hardness of the input instance. The sample complexity is shown to be multiplicatively better depending on the length of rank-ordered feedback available in each subset play. We also give a new fixed-budget best-item algorithm for the \pl\, model along with an error bound. Numerical results of simulations of the algorithms are reported.