We consider the problem of learning a mixture of linear regressions (MLRs). An MLR is specified by $k$ nonnegative mixing weights $p_1, \ldots, p_k$ summing to $1$, and $k$ unknown regressors $w_1,...,w_k\in\mathbb{R}^d$. A sample from the MLR is drawn by sampling $i$ with probability $p_i$, then outputting $(x, y)$ where $y = \langle x, w_i \rangle + \eta$, where $\eta\sim\mathcal{N}(0,\varsigma^2)$ for noise rate $\varsigma$. Mixtures of linear regressions are a popular generative model and have been studied extensively in machine learning and theoretical computer science. However, all previous algorithms for learning the parameters of an MLR require running time and sample complexity scaling exponentially with $k$. In this paper, we give the first algorithm for learning an MLR that runs in time which is sub-exponential in $k$. Specifically, we give an algorithm which runs in time $\widetilde{O}(d)\cdot\exp(\widetilde{O}(\sqrt{k}))$ and outputs the parameters of the MLR to high accuracy, even in the presence of nontrivial regression noise. We demonstrate a new method that we call "Fourier moment descent" which uses univariate density estimation and low-degree moments of the Fourier transform of suitable univariate projections of the MLR to iteratively refine our estimate of the parameters. To the best of our knowledge, these techniques have never been used in the context of high dimensional distribution learning, and may be of independent interest. We also show that our techniques can be used to give a sub-exponential time algorithm for learning mixtures of hyperplanes, a natural hard instance of the subspace clustering problem.

In this paper, we explore the limitations of PCA as a dimension reduction technique and study its extension, projection pursuit (PP), which is a broad class of linear dimension reduction methods. PCA is a popular dimension reduction technique commonly applied to scRNA sequencing data. Despite of huge success in practice, we will illustrate three drawbacks of PCA. It is well known that the eigenvalues of sample covariance matrix is not consistent in high dimensional cases. Every principal component is uncorrelated with each other but not independent.

We consider the problem of clustering datasets in the presence of arbitrary outliers. Traditional clustering algorithms such as k-means and spectral clustering are known to perform poorly for datasets contaminated with even a small number of outliers. In this paper, we develop a provably robust spectral clustering algorithm that applies a simple rounding scheme to denoise a Gaussian kernel matrix built from the data points, and uses vanilla spectral clustering to recover the cluster labels of data points. We analyze the performance of our algorithm under the assumption that the "good" inlier data points are generated from a mixture of sub-gaussians, while the "noisy" outlier points can come from any arbitrary probability distribution. For this general class of models, we show that the asymptotic mis-classification error decays at an exponential rate in the signal-to-noise ratio, provided the number of outliers are a small fraction of the inlier points. Surprisingly, the derived error bound matches with the best-known bound for semidefinite programs (SDPs) under the same setting without outliers. We conduct extensive experiments on a variety of simulated and real-world datasets to demonstrate that our algorithm is less sensitive to outliers compared to other state-of-the-art algorithms proposed in the literature, in terms of both accuracy as well as scalability.

Support vector machine (SVM) has attracted great attentions for the last two decades due to its extensive applications, and thus numerous optimization models have been proposed. To distinguish all of them, in this paper, we introduce a new model equipped with an $L_{0/1}$ soft-margin loss (dubbed as $L_{0/1}$-SVM) which well captures the nature of the binary classification. Many of the existing convex/non-convex soft-margin losses can be viewed as a surrogate of the $L_{0/1}$ soft-margin loss. Despite the discrete nature of $L_{0/1}$, we manage to establish the existence of global minimizer of the new model as well as revealing the relationship among its minimizers and KKT/P-stationary points. These theoretical properties allow us to take advantage of the alternating direction method of multipliers. In addition, the $L_{0/1}$-support vector operator is introduced as a filter to prevent outliers from being support vectors during the training process. Hence, the method is expected to be relatively robust. Finally, numerical experiments demonstrate that our proposed method generates better performance in terms of much shorter computational time with much fewer number of support vectors when against with some other leading methods in areas of SVM. When the data size gets bigger, its advantage becomes more evident.

Estimating arbitrary quantities of interest (QoIs) that are non-linear operators of complex, expensive-to-evaluate, black-box functions is a challenging problem due to missing domain knowledge and finite budgets. Bayesian optimal design of experiments (BODE) is a family of methods that identify an optimal design of experiments (DOE) under different contexts, using only in a limited number of function evaluations. Under BODE methods, sequential design of experiments (SDOE) accomplishes this task by selecting an optimal sequence of experiments while using data-driven probabilistic surrogate models instead of the expensive black-box function. Probabilistic predictions from the surrogate model are used to define an information acquisition function (IAF) which quantifies the marginal value contributed or the expected information gained by a hypothetical experiment. The next experiment is selected by maximizing the IAF. A generally applicable IAF is the expected information gain (EIG) about a QoI as captured by the expectation of the Kullback-Leibler divergence between the predictive distribution of the QoI after doing a hypothetical experiment and the current predictive distribution about the same QoI. We model the underlying information source as a fully-Bayesian, non-stationary Gaussian process (FBNSGP), and derive an approximation of the information gain of a hypothetical experiment about an arbitrary QoI conditional on the hyper-parameters The EIG about the same QoI is estimated by sample averages to integrate over the posterior of the hyper-parameters and the potential experimental outcomes. We demonstrate the performance of our method in four numerical examples and a practical engineering problem of steel wire manufacturing. The method is compared to two classic SDOE methods: random sampling and uncertainty sampling.

Abstract-- VLSI mask optimization is one of the most critical stages in manufacturability aware design, which is costly due to the complicated mask optimization and lithography simulation. Recent researches have shown prominent advantages of machine learning techniques dealing with complicated and big data problems, which bring potential of dedicated machine learning solution for DFM problems and facilitate the VLSI design cycle. In this paper, we focus on a heterogeneous OPC framework that assists mask layout optimization. Preliminary results show the efficiency and effectiveness of proposed frameworks that have the potential to be alternatives to existing EDA solutions. I Introduction VLSI mask optimization is one of the most critical stages in manufacturability aware design, which is costly due to the complicated mask optimization and lithography simulation. Recent studies have shown prominent advantages of machine learning techniques dealing with complicated and big data problems, which bring the potential of dedicated machine learning solution for DFM problems and facilitate the VLSI design cycle [1, 2].

Multi-view subspace clustering has been applied to applications such as image processing and video surveillance, and has attracted increasing attention. Most existing methods learn view-specific self-representation matrices, and construct a combined affinity matrix from multiple views. The affinity construction process is time-consuming, and the combined affinity matrix is not guaranteed to reflect the whole true subspace structure. To overcome these issues, the Latent Complete Row Space Recovery (LCRSR) method is proposed. Concretely, LCRSR is based on the assumption that the multi-view observations are generated from an underlying latent representation, which is further assumed to collect the authentic samples drawn exactly from multiple subspaces. LCRSR is able to recover the row space of the latent representation, which not only carries complete information from multiple views but also determines the subspace membership under certain conditions. LCRSR does not involve the graph construction procedure and is solved with an efficient and convergent algorithm, thereby being more scalable to large-scale datasets. The effectiveness and efficiency of LCRSR are validated by clustering various kinds of multi-view data and illustrated in the background subtraction task.

In this expository note we describe a surprising phenomenon in overparameterized linear regression, where the dimension exceeds the number of samples: there is a regime where the test risk of the estimator found by gradient descent increases with additional samples. In other words, more data actually hurts the estimator. This behavior is implicit in a recent line of theoretical works analyzing "double-descent" phenomenon in linear models. In this note, we isolate and understand this behavior in an extremely simple setting: linear regression with isotropic Gaussian covariates. In particular, this occurs due to an unconventional type of bias-variance tradeoff in the overparameterized regime: the bias decreases with more samples, but variance increases.

The method of covariate adjustment is often used for estimation of population average treatment effects in observational studies. Graphical rules for determining all valid covariate adjustment sets from an assumed causal graphical model are well known. Restricting attention to causal linear models, a recent article derived two novel graphical criteria: one to compare the asymptotic variance of linear regression treatment effect estimators that control for certain distinct adjustment sets and another to identify the optimal adjustment set that yields the least squares treatment effect estimator with the smallest asymptotic variance among consistent adjusted least squares estimators. In this paper we show that the same graphical criteria can be used in non-parametric causal graphical models when treatment effects are estimated by contrasts involving non-parametrically adjusted estimators of the interventional means. We also provide a graphical criterion for determining the optimal adjustment set among the minimal adjustment sets, which is valid for both linear and non-parametric estimators. We provide a new graphical criterion for comparing time dependent adjustment sets, that is, sets comprised by covariates that adjust for future treatments and that are themselves affected by earlier treatments. We show by example that uniformly optimal time dependent adjustment sets do not always exist. In addition, for point interventions, we provide a sound and complete graphical criterion for determining when a non-parametric optimally adjusted estimator of an interventional mean, or of a contrast of interventional means, is as efficient as an efficient estimator of the same parameter that exploits the information in the conditional independencies encoded in the non-parametric causal graphical model.

We propose Disentanglement based Active Learning (DAL), a new active learning technique based on query synthesis which leverages the concept of disentanglement. Instead of requesting labels from the human oracle, our method automatically labels majority of the datapoints, thus drastically reducing the human labelling budget in active learning. The proposed method uses Information Maximizing Generative Adversarial Nets (InfoGAN) to achieve the task where the active learner provides a feedback on the generation of InfoGAN based on which decision is taken about the datapoints to be queried. Results on two benchmark datasets demonstrate that DAL is able to achieve nearly fully supervised accuracy with fairly less labelling budget compared to existing active learning approaches.