A Neural Dirichlet Process Mixture Model for Task-Free Continual Learning

Abstract

Despite the growing interest in continual learning, most of its contemporary works have been studied in a rather restricted setting where tasks are clearly distinguishable, and task boundaries are known during training. However, if our goal is to develop an algorithm that learns as humans do, this setting is far from realistic, and it is essential to develop a methodology that works in a task-free manner. Meanwhile, among several branches of continual learning, expansion-based methods have the advantage of eliminating catastrophic forgetting by allocating new resources to learn new data. In this work, we propose an expansion-based approach for task-free continual learning. Our model, named Continual Neural Dirichlet Process Mixture (CN-DPM), consists of a set of neural network experts that are in charge of a subset of the data. CN-DPM expands the number of experts in a principled way under the Bayesian nonparametric framework. With extensive experiments, we show that our model successfully performs task-free continual learning for both discriminative and generative tasks such as image classification and image generation.

Accepted as a conference paper at ICLR 2020

References in corpus (10)

• Auto-Encoding Variational Bayes
• Overcoming catastrophic forgetting in neural networks
• Pixel Recurrent Neural Networks
• Continual Learning Through Synaptic Intelligence
• Efficient Lifelong Learning with A-GEM
• Progress & Compress: A scalable framework for continual learning
• Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
• Nonlinear Models Using Dirichlet Process Mixtures
• Re-evaluating Continual Learning Scenarios: A Categorization and Case for Strong Baselines
• Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting

Cited by in corpus (28)

• Continual Learning and Catastrophic Forgetting
• Recent Advances of Continual Learning in Computer Vision: An Overview
• Meta-Consolidation for Continual Learning
• A Dirichlet Process Mixture of Robust Task Models for Scalable Lifelong Reinforcement Learning
• Lifelong Adaptive Machine Learning for Sensor-based Human Activity Recognition Using Prototypical Networks
• Online Class-Incremental Continual Learning with Adversarial Shapley Value
• Hierarchically Organized Latent Modules for Exploratory Search in Morphogenetic Systems
• Enhancing Consistency and Mitigating Bias: A Data Replay Approach for Incremental Learning
• Hierarchical Indian Buffet Neural Networks for Bayesian Continual Learning
• Encoders and Ensembles for Task-Free Continual Learning
• Supervised Contrastive Replay: Revisiting the Nearest Class Mean Classifier in Online Class-Incremental Continual Learning
• Imbalanced Continual Learning with Partitioning Reservoir Sampling
• Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors
• Hypernetworks for Continual Semi-Supervised Learning
• Memory-Efficient Semi-Supervised Continual Learning: The World is its Own Replay Buffer
• Evaluating Online Continual Learning with CALM
• DrNAS: Dirichlet Neural Architecture Search
• A Probabilistic Framework for Adapting to Changing and Recurring Concepts in Data Streams
• Online Continual Learning on Class Incremental Blurry Task Configuration with Anytime Inference
• Posterior Meta-Replay for Continual Learning
• Neuromodulated Neural Architectures with Local Error Signals for Memory-Constrained Online Continual Learning
• Detecting and Adapting to Irregular Distribution Shifts in Bayesian Online Learning
• Graph-Based Continual Learning
• Online Learning of a Probabilistic and Adaptive Scene Representation
• Task-Agnostic Online Reinforcement Learning with an Infinite Mixture of Gaussian Processes
• Unsupervised Class-Incremental Learning Through Confusion
• Continual Novelty Detection
• Online Continual Learning in Image Classification: An Empirical Survey