This paper presents a hierarchical framework for Deep Reinforcement Learning that acquires motor skills for a variety of push recovery and balancing behaviors, i.e., ankle, hip, foot tilting, and stepping strategies. The policy is trained in a physics simulator with realistic setting of robot model and low-level impedance control that are easy to transfer the learned skills to real robots. The advantage over traditional methods is the integration of high-level planner and feedback control all in one single coherent policy network, which is generic for learning versatile balancing and recovery motions against unknown perturbations at arbitrary locations (e.g., legs, torso). Furthermore, the proposed framework allows the policy to be learned quickly by many state-of-the-art learning algorithms. By comparing our learned results to studies of preprogrammed, special-purpose controllers in the literature, self-learned skills are comparable in terms of disturbance rejection but with additional advantages of producing a wide range of adaptive, versatile and robust behaviors.

Convolutional neural networks are among the most successful architectures in deep learning. This success is at least partially attributable to the efficacy of spatial invariance as an inductive bias. Locally connected layers, which differ from convolutional layers in their lack of spatial invariance, usually perform poorly in practice. However, these observations still leave open the possibility that some degree of relaxation of spatial invariance may yield a better inductive bias than either convolution or local connectivity. To test this hypothesis, we design a method to relax the spatial invariance of a network layer in a controlled manner. In particular, we create a \textit{low-rank} locally connected layer, where the filter bank applied at each position is constructed as a linear combination of basis set of filter banks. By varying the number of filter banks in the basis set, we can control the degree of departure from spatial invariance. In our experiments, we find that relaxing spatial invariance improves classification accuracy over both convolution and locally connected layers across MNIST, CIFAR-10, and CelebA datasets. These results suggest that spatial invariance in convolution layers may be overly restrictive.

The process of neural network pruning with suitable fine-tuning and retraining can yield networks with considerably fewer parameters than the original with comparable degrees of accuracy. Typically, pruning methods require large, pre-trained networks as a starting point from which they perform a time-intensive iterative pruning and retraining algorithm. We propose a novel pruning in-training method that prunes a network real-time during training, reducing the overall training time to achieve an optimal compressed network. To do so, we introduce an activation density based analysis that identifies the optimal relative sizing or compression for each layer of the network. Our method removes the need for pre-training and is architecture agnostic, allowing it to be employed on a wide variety of systems. For VGG-19 and ResNet18 on CIFAR-10, CIFAR-100, and TinyImageNet, we obtain exceedingly sparse networks (up to 200x reduction in parameters and >60x reduction in inference compute operations in the best case) with comparable accuracies (up to 2%-3% loss with respect to the baseline network). By reducing the network size periodically during training, we achieve total training times that are shorter than those of previously proposed pruning methods. Furthermore, training compressed networks at different epochs with our proposed method yields considerable reduction in training compute complexity (1.6x -3.2x lower) at near iso-accuracy as compared to a baseline network trained entirely from scratch.

Drug discovery projects entail cycles of design, synthesis, and testing that yield a series of chemically related small molecules whose properties, such as binding affinity to a given target protein, are progressively tailored to a particular drug discovery goal. The use of deep learning technologies could augment the typical practice of using human intuition in the design cycle, and thereby expedite drug discovery projects. Here we present DESMILES, a deep neural network model that advances the state of the art in machine learning approaches to molecular design. We applied DESMILES to a previously published benchmark that assesses the ability of a method to modify input molecules to inhibit the dopamine receptor D2, and DESMILES yielded a 77% lower failure rate compared to state-of-the-art models. To explain the ability of DESMILES to hone molecular properties, we visualize a layer of the DESMILES network, and further demonstrate this ability by using DESMILES to tailor the same molecules used in the D2 benchmark test to dock more potently against seven different receptors.

The notion of task similarity is at the core of various machine learning paradigms, such as domain adaptation and meta-learning. Current methods to quantify it are often heuristic, make strong assumptions on the label sets across the tasks, and many are architecture-dependent, relying on task-specific optimal parameters (e.g., require training a model on each dataset). In this work we propose an alternative notion of distance between datasets that (i) is model-agnostic, (ii) does not involve training, (iii) can compare datasets even if their label sets are completely disjoint and (iv) has solid theoretical footing. This distance relies on optimal transport, which provides it with rich geometry awareness, interpretable correspondences and well-understood properties. Our results show that this novel distance provides meaningful comparison of datasets, and correlates well with transfer learning hardness across various experimental settings and datasets.

Stochastic gradient Markov chain Monte Carlo (MCMC) algorithms have received much attention in Bayesian computing for big data problems, but they are only applicable to a small class of problems for which the parameter space has a fixed dimension and the log-posterior density is differentiable with respect to the parameters. This paper proposes an extended stochastic gradient MCMC lgoriathm which, by introducing appropriate latent variables, can be applied to more general large-scale Bayesian computing problems, such as those involving dimension jumping and missing data. Numerical studies show that the proposed algorithm is highly scalable and much more efficient than traditional MCMC algorithms. The proposed algorithms have much alleviated the pain of Bayesian methods in big data computing.

Group invariant and equivariant Multilayer Perceptrons (MLP), also known as Equivariant Networks, have achieved remarkable success in learning on a variety of data structures, such as sequences, images, sets, and graphs. Using tools from group theory, this paper proves the universality of a broad class of equivariant MLPs with a single hidden layer. In particular, it is shown that having a hidden layer on which the group acts regularly is sufficient for universal equivariance. Next, Burnside's table of marks is used to decompose product spaces. It is shown that the product of two G-sets always contains an orbit larger than the input orbits. Therefore high order hidden layers inevitably contain a regular orbit, leading to the universality of the corresponding MLP. It is shown that with an order larger than the logarithm of the size of the stabilizer group, a high-order equivariant MLP is equivariant universal.

Feature selection from a large number of covariates (aka features) in a regression analysis remains a challenge in data science, especially in terms of its potential of scaling to ever-enlarging data and finding a group of scientifically meaningful features. For example, to develop new, responsive drug targets for ovarian cancer, the actual false discovery rate (FDR) of a practical feature selection procedure must also match the target FDR. The popular approach to feature selection, when true features are sparse, is to use a penalized likelihood or a shrinkage estimation, such as a LASSO, SCAD, Elastic Net, or MCP procedure (call them benchmark procedures). We present a different approach using a new subsampling method, called the Subsampling Winner algorithm (SWA). The central idea of SWA is analogous to that used for the selection of US national merit scholars. SWA uses a "base procedure" to analyze each of the subsamples, computes the scores of all features according to the performance of each feature from all subsample analyses, obtains the "semifinalist" based on the resulting scores, and then determines the "finalists," i.e., the most important features. Due to its subsampling nature, SWA can scale to data of any dimension in principle. The SWA also has the best-controlled actual FDR in comparison with the benchmark procedures and the randomForest, while having a competitive true-feature discovery rate. We also suggest practical add-on strategies to SWA with or without a penalized benchmark procedure to further assure the chance of "true" discovery. Our application of SWA to the ovarian serous cystadenocarcinoma specimens from the Broad Institute revealed functionally important genes and pathways, which we verified by additional genomics tools. This second-stage investigation is essential in the current discussion of the proper use of P-values.

We investigate the problem of algorithmic fairness in the case where sensitive and non-sensitive features are available and one aims to generate new, `oblivious', features that closely approximate the non-sensitive features, and are only minimally dependent on the sensitive ones. We study this question in the context of kernel methods. We analyze a relaxed version of the Maximum Mean Discrepancy criterion which does not guarantee full independence but makes the optimization problem tractable. We derive a closed-form solution for this relaxed optimization problem and complement the result with a study of the dependencies between the newly generated features and the sensitive ones. Our key ingredient for generating such oblivious features is a Hilbert-space-valued conditional expectation, which needs to be estimated from data. We propose a plug-in approach and demonstrate how the estimation errors can be controlled. Our theoretical results are accompanied by experimental evaluations.

Can meta-learning discover generic ways of processing time-series (TS) from a diverse dataset so as to greatly improve generalization on new TS coming from different datasets? This work provides positive evidence to demonstrate this using a broad meta-learning framework which we show subsumes many existing meta-learning algorithms as specific cases. We further identify via theoretical analysis the meta-learning adaptation mechanisms within N-BEATS, a recent neural TS forecasting model. Our meta-learning theory predicts that N-BEATS iteratively generates a subset of its task-specific parameters based on a given TS input, thus gradually expanding the expressive power of the architecture on-the-fly. Our empirical results emphasize the importance of meta-learning for successful zero-shot forecasting to new sources of TS, supporting the claim that it is viable to train a neural network on a source TS dataset and deploy it on a different target TS dataset without retraining, resulting in performance that is at least as good as that of state-of-practice univariate forecasting models.