We propose random hinge forests, a simple, efficient, and novel variant of decision forests. Importantly, random hinge forests can be readily incorporated as a general component within arbitrary computation graphs that are optimized end-to-end with stochastic gradient descent or variants thereof. We derive random hinge forest and ferns, focusing on their sparse and efficient nature, their min-max margin property, strategies to initialize them for arbitrary network architectures, and the class of optimizers most suitable for optimizing random hinge forest. The performance and versatility of random hinge forests are demonstrated by experiments incorporating a variety of of small and large UCI machine learning data sets and also ones involving the MNIST, Letter, and USPS image datasets. We compare random hinge forests with random forests and the more recent backpropagating deep neural decision forests.

Accelerated gradient (AG) methods are breakthroughs in convex optimization, improving the convergence rate of the gradient descent method for optimization with smooth functions. However, the analysis of AG methods for non-convex optimization is still limited. It remains an open question whether AG methods from convex optimization can accelerate the convergence of the gradient descent method for finding local minimum of non-convex optimization problems. This paper provides an affirmative answer to this question. In particular, we analyze two renowned variants of AG methods (namely Polyak's Heavy Ball method and Nesterov's Accelerated Gradient method) for extracting the negative curvature from random noise, which is central to escaping from saddle points. By leveraging the proposed AG methods for extracting the negative curvature, we present a new AG algorithm with double loops for non-convex optimization~\footnote{this is in contrast to a single-loop AG algorithm proposed in a recent manuscript~\citep{AGNON}, which directly analyzed the Nesterov's AG method for non-convex optimization and appeared online on November 29, 2017. However, we emphasize that our work is an independent work, which is inspired by our earlier work~\citep{NEON17} and is based on a different novel analysis.}, which converges to second-order stationary point $\x$ such that $\|\nabla f(\x)\|\leq \epsilon$ and $\nabla^2 f(\x)\geq -\sqrt{\epsilon} I$ with $\widetilde O(1/\epsilon^{1.75})$ iteration complexity, improving that of gradient descent method by a factor of $\epsilon^{-0.25}$ and matching the best iteration complexity of second-order Hessian-free methods for non-convex optimization.

PCA is one of the most widely used dimension reduction techniques. A related easier problem is "subspace learning" or "subspace estimation". Given relatively clean data, both are easily solved via singular value decomposition (SVD). The problem of subspace learning or PCA in the presence of outliers is called robust subspace learning or robust PCA (RPCA). For long data sequences, if one tries to use a single lower dimensional subspace to represent the data, the required subspace dimension may end up being quite large. For such data, a better model is to assume that it lies in a low-dimensional subspace that can change over time, albeit gradually. The problem of tracking such data (and the subspaces) while being robust to outliers is called robust subspace tracking (RST). This article provides a magazine-style overview of the entire field of robust subspace learning and tracking. In particular solutions for three problems are discussed in detail: RPCA via sparse+low-rank matrix decomposition (S+LR), RST via S+LR, and "robust subspace recovery". This assumes that an entire data vector is either an outlier or an inlier. The S+LR formulation instead assumes that outliers occur on only a few data vector indices and hence are well modeled as sparse corruptions.

Although many contextual bandit algorithms have similar theoretical guarantees, the characteristics of real-world applications oftentimes result in large performance dissimilarities across algorithms. We study a consideration for the exploration vs. exploitation framework that does not arise in non-contextual bandits: the way exploration is conducted in the present may affect the bias and variance in the potential outcome model estimation in subsequent stages of learning. We show that contextual bandit algorithms are sensitive to the estimation method of the outcome model as well as the exploration method used, particularly in the presence of rich heterogeneity or complex outcome models, which can lead to difficult estimation problems along the path of learning. We propose new contextual bandit designs, combining parametric and non-parametric statistical estimation methods with causal inference methods in order to reduce the estimation bias that results from adaptive treatment assignment. We provide empirical evidence that guides the choice among the alternatives in different scenarios, such as prejudice (non-representative user contexts) in the initial training data.

Two classes of methods have been proposed for escaping from saddle points with one using the second-order information carried by the Hessian and the other adding the noise into the first-order information. The existing analysis for algorithms using noise in the first-order information is quite involved and hides the essence of added noise, which hinder further improvements of these algorithms. In this paper, we present a novel perspective of noise-adding technique, i.e., adding the noise into the first-order information can help extract the negative curvature from the Hessian matrix, and provide a formal reasoning of this perspective by analyzing a simple first-order procedure. More importantly, the proposed procedure enables one to design purely first-order stochastic algorithms for escaping from non-degenerate saddle points with a much better time complexity (almost linear time in terms of the problem's dimensionality). In particular, we develop a {\bf first-order stochastic algorithm} based on our new technique and an existing algorithm that only converges to a first-order stationary point to enjoy a time complexity of {$\widetilde O(d/\epsilon^{3.5})$ for finding a nearly second-order stationary point $\bf{x}$ such that $\|\nabla F(bf{x})\|\leq \epsilon$ and $\nabla^2 F(bf{x})\geq -\sqrt{\epsilon}I$ (in high probability), where $F(\cdot)$ denotes the objective function and $d$ is the dimensionality of the problem. To the best of our knowledge, this is the best theoretical result of first-order algorithms for stochastic non-convex optimization, which is even competitive with if not better than existing stochastic algorithms hinging on the second-order information.

Graph convolutional networks (GCNs) are powerful deep neural networks for graph-structured data. However, GCN computes the representation of a node recursively from its neighbors, making the receptive field size grow exponentially with the number of layers. Previous attempts on reducing the receptive field size by subsampling neighbors do not have a convergence guarantee, and their receptive field size per node is still in the order of hundreds. In this paper, we develop control variate based algorithms which allow sampling an arbitrarily small neighbor size. Furthermore, we prove new theoretical guarantee for our algorithms to converge to a local optimum of GCN. Empirical results show that our algorithms enjoy a similar convergence with the exact algorithm using only two neighbors per node. The runtime of our algorithms on a large Reddit dataset is only one seventh of previous neighbor sampling algorithms.

Information bottleneck [IB] is a technique for extracting information in some `input' random variable that is relevant for predicting some different 'output' random variable. IB works by encoding the input in a compressed 'bottleneck variable' from which the output can then be accurately decoded. IB can be difficult to compute in practice, and has been mainly developed for two limited cases: (1) discrete random variables with small state spaces, and (2) continuous random variables that are jointly Gaussian distributed (in which case the encoding and decoding maps are linear). We propose a method to perform IB in more general domains. Our approach can be applied to discrete or continuous inputs and outputs, and allows for nonlinear encoding and decoding maps. The method uses a novel upper bound on the IB objective, derived using a non-parametric estimator of mutual information and a variational approximation. We show how to implement the method using neural networks and gradient-based optimization, and demonstrate its performance on the MNIST dataset.

I am doing some regression analysis. Some of the independent variables are continuous while some are categorical. The dependent variable is continuous. Can you please help me on which regression model should I pick?

Statistics came well before computers. It would be very different if it were the other way around. The stats most people learn in high school or college come from the time when computations were done with pen and paper. "Statistics were constrained by the computational technology available at the time," says Stanford statistics professor Robert Tibshirani. "People use certain methods because that is how it all started and that's what they are used to. It's hard to change it."

We propose a Multi-Instance-Learning (MIL) approach for weakly-supervised learning problems, where a training set is formed by bags (sets of feature vectors or instances) and only labels at bag-level are provided. Specifically, we consider the Multi-Instance Dynamic-Ordinal-Regression (MI-DOR) setting, where the instance labels are naturally represented as ordinal variables and bags are structured as temporal sequences. To this end, we propose Multi-Instance Dynamic Ordinal Random Fields (MI-DORF). In this framework, we treat instance-labels as temporally-dependent latent variables in an Undirected Graphical Model. Different MIL assumptions are modelled via newly introduced high-order potentials relating bag and instance-labels within the energy function of the model. We also extend our framework to address the Partially-Observed MI-DOR problems, where a subset of instance labels are available during training. We show on the tasks of weakly-supervised facial behavior analysis, Facial Action Unit (DISFA dataset) and Pain (UNBC dataset) Intensity estimation, that the proposed framework outperforms alternative learning approaches. Furthermore, we show that MIDORF can be employed to reduce the data annotation efforts in this context by large-scale.