Learning to Rank For Push Notifications Using Pairwise Expected Regret

Listwise ranking losses have been widely studied in recommender systems. However, new paradigms of content consumption present new challenges for ranking methods. In this work we contribute an analysis of learning to rank for personalized mobile push notifications and discuss the unique challenges this presents compared to traditional ranking problems. To address these challenges, we introduce a novel ranking loss based on weighting the pairwise loss between candidates by the expected regret incurred for misordering the pair. We demonstrate that the proposed method can outperform prior methods both in a simulated environment and in a production experiment on a major social network.

Implicit Distributional Reinforcement Learning

To improve the sample efficiency of policy-gradient based reinforcement learning algorithms, we propose implicit distributional actor critic (IDAC) that consists of a distributional critic, built on two deep generator networks (DGNs), and a semi-implicit actor (SIA), powered by a flexible policy distribution. We adopt a distributional perspective on the discounted cumulative return and model it with a state-action-dependent implicit distribution, which is approximated by the DGNs that take state-action pairs and random noises as their input. Moreover, we use the SIA to provide a semi-implicit policy distribution, which mixes the policy parameters with a reparameterizable distribution that is not constrained by an analytic density function. In this way, the policy's marginal distribution is implicit, providing the potential to model complex properties such as covariance structure and skewness, but its parameter and entropy can still be estimated. We incorporate these features with an off-policy algorithm framework to solve problems with continuous action space, and compare IDAC with the state-of-art algorithms on representative OpenAI Gym environments. We observe that IDAC outperforms these baselines for most tasks.

Discrete Action On-Policy Learning with Action-Value Critic

Reinforcement learning (RL) in discrete action space is ubiquitous in real-world applications, but its complexity grows exponentially with the action-space dimension, making it challenging to apply existing on-policy gradient based deep RL algorithms efficiently. To effectively operate in multidimensional discrete action spaces, we construct a critic to estimate action-value functions, apply it on correlated actions, and combine these critic estimated action values to control the variance of gradient estimation. We follow rigorous statistical analysis to design how to generate and combine these correlated actions, and how to sparsify the gradients by shutting down the contributions from certain dimensions. These efforts result in a new discrete action on-policy RL algorithm that empirically outperforms related on-policy algorithms relying on variance control techniques. We demonstrate these properties on OpenAI Gym benchmark tasks, and illustrate how discretizing the action space could benefit the exploration phase and hence facilitate convergence to a better local optimal solution thanks to the flexibility of discrete policy.

Semi-supervised Learning using Adversarial Training with Good and Bad Samples

In this work, we investigate semi-supervised learning (SSL) for image classification using adversarial training. Previous results have illustrated that generative adversarial networks (GANs) can be used for multiple purposes. Triple-GAN, which aims to jointly optimize model components by incorporating three players, generates suitable image-label pairs to compensate for the lack of labeled data in SSL with improved benchmark performance. Conversely, Bad (or complementary) GAN, optimizes generation to produce complementary data-label pairs and force a classifier's decision boundary to lie between data manifolds. Although it generally outperforms Triple-GAN, Bad GAN is highly sensitive to the amount of labeled data used for training. Unifying these two approaches, we present unified-GAN (UGAN), a novel framework that enables a classifier to simultaneously learn from both good and bad samples through adversarial training. We perform extensive experiments on various datasets and demonstrate that UGAN: 1) achieves state-of-the-art performance among other deep generative models, and 2) is robust to variations in the amount of labeled data used for training.

A Unified Framework for Tuning Hyperparameters in Clustering Problems

Selecting hyperparameters for unsupervised learning problems is difficult in general due to the lack of ground truth for validation. However, this issue is prevalent in machine learning, especially in clustering problems with examples including the Lagrange multipliers of penalty terms in semidefinite programming (SDP) relaxations and the bandwidths used for constructing kernel similarity matrices for Spectral Clustering. Despite this, there are not many provable algorithms for tuning these hyperparameters. In this paper, we provide a unified framework with provable guarantees for the above class of problems. We demonstrate our method on two distinct models. First, we show how to tune the hyperparameters in widely used SDP algorithms for community detection in networks. In this case, our method can also be used for model selection. Second, we show the same framework works for choosing the bandwidth for the kernel similarity matrix in Spectral Clustering for subgaussian mixtures under suitable model specification. In a variety of simulation experiments, we show that our framework outperforms other widely used tuning procedures in a broad range of parameter settings.

ARSM: Augment-REINFORCE-Swap-Merge Estimator for Gradient Backpropagation Through Categorical Variables

To address the challenge of backpropagating the gradient through categorical variables, we propose the augment-REINFORCE-swap-merge (ARSM) gradient estimator that is unbiased and has low variance. ARSM first uses variable augmentation, REINFORCE, and Rao-Blackwellization to re-express the gradient as an expectation under the Dirichlet distribution, then uses variable swapping to construct differently expressed but equivalent expectations, and finally shares common random numbers between these expectations to achieve significant variance reduction. Experimental results show ARSM closely resembles the performance of the true gradient for optimization in univariate settings; outperforms existing estimators by a large margin when applied to categorical variational auto-encoders; and provides a "try-and-see self-critic" variance reduction method for discrete-action policy gradient, which removes the need of estimating baselines by generating a random number of pseudo actions and estimating their action-value functions.