Recently, there has been a growing interest in the problem of learning rich implicit models - those from which we can sample, but can not evaluate their density. These models apply some parametric function, such as a deep network, to a base measure, and are learned end-to-end using stochastic optimization. One strategy of devising a loss function is through the statistics of two sample tests - if we can fool a statistical test, the learned distribution should be a good model of the true data. However, not all tests can easily fit into this framework, as they might not be differentiable with respect to the data points, and hence with respect to the parameters of the implicit model. Motivated by this problem, in this paper we show how two such classical tests, the Friedman-Rafsky and k-nearest neighbour tests, can be effectively smoothed using ideas from undirected graphical models - the matrix tree theorem and cardinality potentials. Moreover, as we show experimentally, smoothing can significantly increase the power of the test, which might of of independent interest. Finally, we apply our method to learn implicit models.

The small-ball method was introduced as a way of obtaining a high probability, isomorphic lower bound on the quadratic empirical process, under weak assumptions on the indexing class. The key assumption was that class members satisfy a uniform small-ball estimate, that is, $Pr(|f| \geq \kappa\|f\|_{L_2}) \geq \delta$ for given constants $\kappa$ and $\delta$. Here we extend the small-ball method and obtain a high probability, almost-isometric (rather than isomorphic) lower bound on the quadratic empirical process. The scope of the result is considerably wider than the small-ball method: there is no need for class members to satisfy a uniform small-ball condition, and moreover, motivated by the notion of tournament learning procedures, the result is stable under a `majority vote'. As applications, we study the performance of empirical risk minimization in learning problems involving bounded subsets of $L_p$ that satisfy a Bernstein condition, and of the tournament procedure in problems involving bounded subsets of $L_\infty$.

Stochastic gradient Markov Chain Monte Carlo (SG-MCMC) has been developed as a flexible family of scalable Bayesian sampling algorithms. However, there has been little theoretical analysis of the impact of minibatch size to the algorithm's convergence rate. In this paper, we prove that under a limited computational budget/time, a larger minibatch size leads to a faster decrease of the mean squared error bound (thus the fastest one corresponds to using full gradients), which motivates the necessity of variance reduction in SG-MCMC. Consequently, by borrowing ideas from stochastic optimization, we propose a practical variance-reduction technique for SG-MCMC, that is efficient in both computation and storage. We develop theory to prove that our algorithm induces a faster convergence rate than standard SG-MCMC. A number of large-scale experiments, ranging from Bayesian learning of logistic regression to deep neural networks, validate the theory and demonstrate the superiority of the proposed variance-reduction SG-MCMC framework.

Very frequently, tensor mining is done in an entirely unsupervised way, since ground truth and labels are either very expensive or hard to obtain. Our problem, thus, is: given a potentially very large and sparse tensor, and its R-component decomposition, compute a quality measure for that decomposition. Subsequently, using that quality metric, we would like to identify a "good" number R of components, and ultimately minimize human intervention and trial-and-error fine tuning. This problem is extremely hard. In fact, even computing the rank of a tensor has been shown to be an NPhard problem, in stark contrast to the matrix rank which can be easily computed in polynomial time. Fortunately, there exist heuristics that are able to assist with the above problem and have been shown to work well in practice, in the field of Chemometrics. Such a powerful and intuitive heuristic is the so-called "Core Consistency Diagnostic" [1], which given a tensor and its PARAFAC decomposition, provides a quality measure, which we can in turn use as a proxy of how interpretable our results are.

In essence, people who interact with advanced technology want to be able to trust it appropriately, and then act on that trust. In interpersonal relationships, and otherwise, humans act largely based on trust. For example, a supervisor asks a subordinate to accomplish a task based on several factors that indicate they can trust them to accomplish that task. When consumers make purchases, they do so with trust that the product will perform as promised. Likewise, when using something like an autonomous vehicle, the user must be able to trust it appropriately in order to use it properly. With the rapid advancement of the capabilities of intelligent computing technology to do tasks that were previously assumed to be too complicated for computers, there has been much recent discussion regarding how humans can trust this technology - although the connection to trust is not always made explicit, per se.

Today's artificial assistants are typically prompted to perform tasks through direct, imperative commands such as \emph{Set a timer} or \emph{Pick up the box}. However, to progress toward more natural exchanges between humans and these assistants, it is important to understand the way non-imperative utterances can indirectly elicit action of an addressee. In this paper, we investigate command types in the setting of a grounded, collaborative game. We focus on a less understood family of utterances for eliciting agent action, locatives like \emph{The chair is in the other room}, and demonstrate how these utterances indirectly command in specific game state contexts. Our work shows that models with domain-specific grounding can effectively realize the pragmatic reasoning that is necessary for more robust natural language interaction.

Several weeks ago I wrote a post on Bayesian statistics. I was very interested in the implementation of Bayesian statistics, especially for complex problems which are more easily solved with simulation rather than mathematical manipulation. I wrote the article with a specific audience in mind: namely those that knew the basics of Bayesian statistics, but had no idea how to implement it. As an astute commenter pointed out, in my excitement to implement my Bayesian program, I skimmed over several key points of Bayesian statistics and woefully mis-represented others. Let's fix that now, shall we! Let's talk about the basics of Bayesian statistics, and then move up to simulating them.

We will walk you through all the aspects of machine learning from simple linear regressions to the latest neural networks, and you will learn not only how to use them but also how to build them from scratch. Big part of this path is oriented on Computer Vision(CV), because it's the fastest way to get general knowledge, and the experience from CV can be simply transferred to any ML area. We will use TensorFlow as a ML framework, as it is the most promising and production ready. Learning will be better if you work on theoretical and practical materials at the same time to get practical experience on the learned material. Also if you want to compete with other people solving real life problems I would recommend you to register on Kaggle, as it could be a good addition to your resume.

February 7, 2017 Written by: Susan C. Daffron IBM surveyed more than 600 decision-makers about their cognitive initiatives and 62 percent of respondents stated that the results of their cognitive implementations exceed expectations*. Cognitive services, like those offered byIBM Watson, can help you find out how your customers feel and help you predict what they might do. With Watson, IBM is pioneering the development of models that can tell you about different and often hidden, aspects of an individual. These insights can then be used by an organization to deepen relationships, shape initiatives and drive innovation. REST APIs, like Watson Personality Insights and Watson Emotion Analysis, allow organizations to learn about an individual's: Organizations can now train apps to quickly analyze and interpret large volumes of unstructured sensory data.

SYS-CON Events announced today that StorageCraft Technology Corp, a global leader in backup and disaster recovery, will exhibit at SYS-CON's 20th International Cloud Expo, which will take place on June 6-8, 2017, at the Javits Center in New York City, NY. The StorageCraft family of companies, founded in 2003, provides award-winning backup, disaster recovery, system migration and data protection solutions for servers, desktops and laptops in addition to powerful data analytics. StorageCraft delivers software products that reduce downtime, improve security and stability for systems and data, and lower the total cost of ownership. All major researchers estimate there will be tens of billions devices - computers, smartphones, tablets, and sensors - connected to the Internet by 2020. This number will continue to grow at a rapid pace for the next several decades.