Contributions to Large Scale Bayesian Inference and Adversarial Machine Learning

Abstract

The rampant adoption of ML methodologies has revealed that models are usually adopted to make decisions without taking into account the uncertainties in their predictions. More critically, they can be vulnerable to adversarial examples. Thus, we believe that developing ML systems that take into account predictive uncertainties and are robust against adversarial examples is a must for critical, real-world tasks. We start with a case study in retailing. We propose a robust implementation of the Nerlove-Arrow model using a Bayesian structural time series model. Its Bayesian nature facilitates incorporating prior information reflecting the manager's views, which can be updated with relevant data. However, this case adopted classical Bayesian techniques, such as the Gibbs sampler. Nowadays, the ML landscape is pervaded with neural networks and this chapter also surveys current developments in this sub-field. Then, we tackle the problem of scaling Bayesian inference to complex models and large data regimes. In the first part, we propose a unifying view of two different Bayesian inference algorithms, Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) and Stein Variational Gradient Descent (SVGD), leading to improved and efficient novel sampling schemes. In the second part, we develop a framework to boost the efficiency of Bayesian inference in probabilistic models by embedding a Markov chain sampler within a variational posterior approximation. After that, we present an alternative perspective on adversarial classification based on adversarial risk analysis, and leveraging the scalable Bayesian approaches from chapter 2. In chapter 4 we turn to reinforcement learning, introducing Threatened Markov Decision Processes, showing the benefits of accounting for adversaries in RL while the agent learns.

arXiv commentPhD thesis

References in corpus (50)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Auto-Encoding Variational Bayes
• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• ADADELTA: An Adaptive Learning Rate Method
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Language Models are Few-Shot Learners
• Machine learning and the physical sciences
• A Survey of Deep Learning Techniques for Autonomous Driving
• SGDR: Stochastic Gradient Descent with Warm Restarts
• Pixel Recurrent Neural Networks
• Semi-Supervised Learning with Deep Generative Models
• Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning
• Variational Inference with Normalizing Flows
• Inferring causal impact using Bayesian structural time-series models
• Stochastic Gradient Hamiltonian Monte Carlo
• Security Evaluation of Pattern Classifiers under Attack
• Autonomous Agents Modelling Other Agents: A Comprehensive Survey and Open Problems
• Robust Adversarial Reinforcement Learning
• Markov Chain Monte Carlo and Variational Inference: Bridging the Gap
• #Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning
• A General Framework for the Parametrization of Hierarchical Models
• On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models
• Bayesian Deep Learning and a Probabilistic Perspective of Generalization
• Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks
• Neural Network Renormalization Group
• Deep Learning: A Bayesian Perspective
• AI Safety Gridworlds
• Stein Variational Gradient Descent as Gradient Flow
• Dropout Inference in Bayesian Neural Networks with Alpha-divergences
• Tactics of Adversarial Attack on Deep Reinforcement Learning Agents
• Yes, but Did It Work?: Evaluating Variational Inference
• Mathematics of Deep Learning
• Inference Suboptimality in Variational Autoencoders
• Variational Inference using Implicit Distributions
• NeuTra-lizing Bad Geometry in Hamiltonian Monte Carlo Using Neural Transport
• A Selective Overview of Deep Learning
• Adversarial classification: An adversarial risk analysis approach
• Learning to Draw Samples with Amortized Stein Variational Gradient Descent
• Semi-Implicit Variational Inference
• Stochastic Gradient MCMC with Repulsive Forces
• A Contrastive Divergence for Combining Variational Inference and MCMC
• Unbiased Implicit Variational Inference
• Reinforcement Learning under Threats
• Classifier-agnostic saliency map extraction
• Assessing the effect of advertising expenditures upon sales: a Bayesian structural time series model
• Functional Tensors for Probabilistic Programming
• Opponent Aware Reinforcement Learning
• A Statistician Teaches Deep Learning