While being a vibrant subfield of computer science, machine learning is used for drawing models and methods from statistics, algorithms, computational complexity, control theory and artificial intelligence. It focuses on efficient algorithms for inferring good predictive models from large data sets and is natural candidate for problems arising in HFT – both trade execution & alpha generation. In quantitative finance inference of models of predictive nature using historical data is obviously not new. Some examples include the coefficient estimation for CAPM, Fama and French factors. The granularity of data arising in HFT poses special challenges for machine learning. Often data microstructure at the resolution of individual orders, executions, hidden liquidity and cancellation including lack of understanding of how such granular data relates to actionable circumstances, namely profitably buying or selling shares, optimally executing a large order, etc.

"Winning a data science competition needs skills that are not relevant to a data science job" That explains why I am very good at the job I don't have;)

Artificial intelligence (AI) may sound like science fiction, but it is real, and becoming increasingly important to companies in every sector. The field of artificial intelligence has produced a wide variety of "cognitive technologies" that simulate human reasoning and perceptual skills, giving businesses entirely new capabilities and enabling organizations to break prevailing tradeoffs between speed, cost, and quality. Aimed at a general business audience, this course demystifies artificial intelligence, provides an overview of a wide range of cognitive technologies, and offers a framework to help you understand their business implications. Some experts have called artificial intelligence "more important than anything since the industrial revolution." That makes this course essential for professionals working in business, operations, strategy, IT, and other disciplines.

There are several reasons why everyone isn't using Bayesian methods for regression modeling. One reason is that Bayesian modeling requires more thought: you need pesky things like priors, and you can't assume that if a procedure runs without throwing an error that the answers are valid. A second reason is that MCMC sampling -- the bedrock of practical Bayesian modeling -- can be slow compared to closed-form or MLE procedures. A third reason is that existing Bayesian solutions have either been highly-specialized (and thus inflexible), or have required knowing how to use a generalized tool like BUGS, JAGS, or Stan. This third reason has recently been shattered in the R world by not one but two packages: brms and rstanarm.

In this paper, we compare 5 different nonlinear kernels: min-max, RBF, fRBF (folded RBF), acos, and acos-$\chi^2$, on a wide range of publicly available datasets. The proposed fRBF kernel performs very similarly to the RBF kernel. Both RBF and fRBF kernels require an important tuning parameter ($\gamma$). Interestingly, for a significant portion of the datasets, the min-max kernel outperforms the best-tuned RBF/fRBF kernels. The acos kernel and acos-$\chi^2$ kernel also perform well in general and in some datasets achieve the best accuracies. One crucial issue with the use of nonlinear kernels is the excessive computational and memory cost. These days, one increasingly popular strategy is to linearize the kernels through various randomization algorithms. In our study, the randomization method for the min-max kernel demonstrates excellent performance compared to the randomization methods for other types of nonlinear kernels, measured in terms of the number of nonzero terms in the transformed dataset. Our study provides evidence for supporting the use of the min-max kernel and the corresponding randomized linearization method (i.e., the so-called "0-bit CWS"). Furthermore, the results motivate at least two directions for future research: (i) To develop new (and linearizable) nonlinear kernels for better accuracies; and (ii) To develop better linearization algorithms for improving the current linearization methods for the RBF kernel, the acos kernel, and the acos-$\chi^2$ kernel. One attempt is to combine the min-max kernel with the acos kernel or the acos-$\chi^2$ kernel. The advantages of these two new and tuning-free nonlinear kernels are demonstrated vias our extensive experiments.

If a document is about travel, we may expect that short snippets of the document should also be about travel. We introduce a general framework for incorporating these types of invariances into a discriminative classifier. The framework imagines data as being drawn from a slice of a Lévy process. If we slice the Lévy process at an earlier point in time, we obtain additional pseudo-examples, which can be used to train the classifier. We show that this scheme has two desirable properties: it preserves the Bayes decision boundary, and it is equivalent to fitting a generative model in the limit where we rewind time back to 0. Our construction captures popular schemes such as Gaussian feature noising and dropout training, as well as admitting new generalizations. Black-box discriminative classifiers such as logistic regression, neural networks, and SVMs are the go-to solution in machine learning: they are simple to apply and often perform well. However, an expert may have additional knowledge to exploit, often taking the form of a certain family of transformations that should usually leave labels fixed. For example, in object recognition, an image of a cat rotated, translated, and peppered with a small amount of noise is probably still a cat.

Crowdsourcing enables researchers to conduct social experiments on a heterogenous set of participants and at a lower economic cost than conventional laboratory studies. For example, researchers can harness internet users to conduct user studies on their personal computers. Among various approaches to conduct subjective tests, pairwise comparisons are expected to yield more reliable results. However, in crowdsourced studies, the individuals performing the ratings are diverse compared to more controlled settings, which is difficult to control for using traditional experimental designs; researchers have recently proposed several randomized methods to conduct user studies [1, 2, 3], which accommodate incomplete and imbalanced data. HodgeRank, as an application of combinatorial Hodge theory to the preference or rank aggregation problem from pairwise comparison data, possibly being incomplete and imbalanced, was first introduced by [4], and inspired a series of studies in statistical ranking [5, 6, 7, 8]. Hodge theory has also found applications in game theory [9] and computer vision [10, 11], in addition to traditional applications in fluid mechanics [12] etc. HodgeRank formulates the ranking problem in terms of the discrete Hodge decomposition of the pairwise data and shows that it can be decomposed into three orthogonal components: a gradient flow representing a global rating (optimal in the L

Having a regression model, we are interested in finding two-sided intervals that are guaranteed to contain at least a desired proportion of the conditional distribution of the response variable given a specific combination of predictors. We name such intervals predictive intervals. This work presents a new method to find two-sided predictive intervals for non-parametric least squares regression without the homoscedasticity assumption. Our predictive intervals are built by using tolerance intervals on prediction errors in the query point's neighborhood. We proposed a predictive interval model test and we also used it as a constraint in our hyper-parameter tuning algorithm. This gives an algorithm that finds the smallest reliable predictive intervals for a given dataset. We also introduce a measure for comparing different interval prediction methods yielding intervals having different size and coverage. These experiments show that our methods are more reliable, effective and precise than other interval prediction methods.

Independent component analysis (ICA), as an approach to the blind source-separation (BSS) problem, has become the de-facto standard in many medical imaging settings. Despite successes and a large ongoing research effort, the limitation of ICA to square linear transformations have not been overcome, so that general INFOMAX is still far from being realized. As an alternative, we present feature analysis in medical imaging as a problem solved by Helmholtz machines, which include dimensionality reduction and reconstruction of the raw data under the same objective, and which recently have overcome major difficulties in inference and learning with deep and nonlinear configurations. We demonstrate one approach to training Helmholtz machines, variational auto-encoders (VAE), as a viable approach toward feature extraction with magnetic resonance imaging (MRI) data.

The stochastic block model is one of the oldest and most ubiquitous models for studying clustering and community detection. In an exciting sequence of developments, motivated by deep but non-rigorous ideas from statistical physics, Decelle et al. conjectured a sharp threshold for when community detection is possible in the sparse regime. Mossel, Neeman and Sly and Massoulie proved the conjecture and gave matching algorithms and lower bounds. Here we revisit the stochastic block model from the perspective of semirandom models where we allow an adversary to make `helpful' changes that strengthen ties within each community and break ties between them. We show a surprising result that these `helpful' changes can shift the information-theoretic threshold, making the community detection problem strictly harder. We complement this by showing that an algorithm based on semidefinite programming (which was known to get close to the threshold) continues to work in the semirandom model (even for partial recovery). This suggests that algorithms based on semidefinite programming are robust in ways that any algorithm meeting the information-theoretic threshold cannot be. These results point to an interesting new direction: Can we find robust, semirandom analogues to some of the classical, average-case thresholds in statistics? We also explore this question in the broadcast tree model, and we show that the viewpoint of semirandom models can help explain why some algorithms are preferred to others in practice, in spite of the gaps in their statistical performance on random models.