The former belongs to the realm of explanatory models, the latter to the realm of predictive analytics. Explanatory models, often involving linear regression, are concerned with explaining a given phenomenon and finding causal relationships between an output (dependent) variable, and a host, often very few, input (independent) variables. The objective is to find a good regression model that fits the data very well which meets the underlying assumption of linear regression. The emphasis here is on hypothesis testing, p-values, confidence intervals,…Once a good model is found, one can use it for estimating the value of the output variable for given values of the input variables. It is OK to estimate an output value based on interpolation, but one must use extreme caution in estimating output values based on extrapolation because the regression model is an explanatory model, not a predictive one.

We consider the problem of estimating the class prior in an unlabeled dataset. Under the assumption that an additional labeled dataset is available, the class prior can be estimated by fitting a mixture of class-wise data distributions to the unlabeled data distribution. However, in practice, such an additional labeled dataset is often not available. In this paper, we show that, with additional samples coming only from the positive class, the class prior of the unlabeled dataset can be estimated correctly. Our key idea is to use properly penalized divergences for model fitting to cancel the error caused by the absence of negative samples. We further show that the use of the penalized $L_1$-distance gives a computationally efficient algorithm with an analytic solution. The consistency, stability, and estimation error are theoretically analyzed. Finally, we experimentally demonstrate the usefulness of the proposed method.

We focus in this work on the estimation of the first $k$ eigenvectors of any graph Laplacian using filtering of Gaussian random signals. We prove that we only need $k$ such signals to be able to exactly recover as many of the smallest eigenvectors, regardless of the number of nodes in the graph. In addition, we address key issues in implementing the theoretical concepts in practice using accurate approximated methods. We also propose fast algorithms both for eigenspace approximation and for the determination of the $k$th smallest eigenvalue $\lambda_k$. The latter proves to be extremely efficient under the assumption of locally uniform distribution of the eigenvalue over the spectrum. Finally, we present experiments which show the validity of our method in practice and compare it to state-of-the-art methods for clustering and visualization both on synthetic small-scale datasets and larger real-world problems of millions of nodes. We show that our method allows a better scaling with the number of nodes than all previous methods while achieving an almost perfect reconstruction of the eigenspace formed by the first $k$ eigenvectors.

We investigate the problem of testing whether $d$ random variables, which may or may not be continuous, are jointly (or mutually) independent. Our method builds on ideas of the two variable Hilbert-Schmidt independence criterion (HSIC) but allows for an arbitrary number of variables. We embed the $d$-dimensional joint distribution and the product of the marginals into a reproducing kernel Hilbert space and define the $d$-variable Hilbert-Schmidt independence criterion (dHSIC) as the squared distance between the embeddings. In the population case, the value of dHSIC is zero if and only if the $d$ variables are jointly independent, as long as the kernel is characteristic. Based on an empirical estimate of dHSIC, we define three different non-parametric hypothesis tests: a permutation test, a bootstrap test and a test based on a Gamma approximation. We prove that the permutation test achieves the significance level and that the bootstrap test achieves pointwise asymptotic significance level as well as pointwise asymptotic consistency (i.e., it is able to detect any type of fixed dependence in the large sample limit). The Gamma approximation does not come with these guarantees; however, it is computationally very fast and for small $d$, it performs well in practice. Finally, we apply the test to a problem in causal discovery.

The amount of data available in the world is growing faster than our ability to deal with it. However, if we take advantage of the internal \emph{structure}, data may become much smaller for machine learning purposes. In this paper we focus on one of the fundamental machine learning tasks, empirical risk minimization (ERM), and provide faster algorithms with the help from the clustering structure of the data. We introduce a simple notion of raw clustering that can be efficiently computed from the data, and propose two algorithms based on clustering information. Our accelerated algorithm ClusterACDM is built on a novel Haar transformation applied to the dual space of the ERM problem, and our variance-reduction based algorithm ClusterSVRG introduces a new gradient estimator using clustering. Our algorithms outperform their classical counterparts ACDM and SVRG respectively.

Decision making with partial feedback, motivated by applications including personalized medicine [22] and content recommendation [17], is receiving increasing attention from the machine learning community. These problems are formally modeled as learning from bandit feedback, where a learner repeatedly takes an action and observes a reward for the action, with the goal of maximizing reward. While bandit learning captures many problems of interest, several applications have additional structure: the action is combinatorial in nature and more detailed feedback is provided. For example, in internet applications, we often recommend sets of items and record information about the user's interaction with each individual item (e.g., click). This additional feedback is unhelpful unless it relates to the overall reward (e.g., number of clicks), and, as in previous work, we assume a linear relationship. This interaction is known as the semibandit feedback model. Typical bandit and semibandit algorithms achieve reward that is competitive with the single best fixed action, i.e., the best medical treatment or the most popular news article for everyone. This is often inadequate for recommendation applications: while the most popular articles may get some clicks, personalizing content to the users is much more effective.

An artificial intelligence method developed by University College London computer scientists and associates has predicted the judicial decisions of the European Court of Human Rights (ECtHR) with 79% accuracy, according to a paper published Monday, Oct. 24 in PeerJ Computer Science. The method is the first to predict the outcomes of a major international court by automatically analyzing case text using a machine-learning algorithm.* "We don't see AI replacing judges or lawyers," said Nikolaos Aletras, who led the study at UCL Computer Science, "but we think they'd find it useful for rapidly identifying patterns in cases that lead to certain outcomes. It could also be a valuable tool for highlighting which cases are most likely to be violations of the European Convention on Human Rights." In developing the method, the team found that judgments by the ECtHR are highly correlated to non-legal (real-world) facts, rather than direct legal arguments, suggesting that judges of the Court are, in the jargon of legal theory, "realists" rather than "formalists."

Machine learning is often thought to be too complicated for everyday development tasks. I have always felt like we can benefit from using machine learning for simple tasks that we do regularly. At Mailgun, we work with e-mail and as part of our offering, we parse HTML quotations. This allows a user to grab the latest reply instead of the entire conversation, which is returned as part of our webhook response. For those of you who don't know, here's what parsing HTML from the public Internet looks like: Changing the parsing library can help, but it won't solve the issue completely because every library has its limitations.

This is an overview (with links) to a 5-part series on introductory machine learning. The set of tutorials is comprehensive, yet succinct, covering many important topics in the field (and beyond). Machine learning is a very hot topic for many key reasons, and because it provides the ability to automatically obtain deep insights, recognize unknown patterns, and create high performing predictive models from data, all without requiring explicit programming instructions. This is a summary (with links) to an article series that's intended to be a comprehensive, in-depth guide to machine learning, and should be useful to everyone from business executives to machine learning practitioners. It covers virtually all aspects of machine learning (and many related fields) at a high level, and should serve as a sufficient introduction or reference to the terminology, concepts, tools, considerations, and techniques in the field.

What has changed in Machine Learning in the past 25 years? You may not care about this question. You may even not realize that Machine Learning as a technical and scientific field is older than 25 years. But I do care about this question. I care because I got a PhD in Machine Learning in 1990.