Task-agnostic Continual Learning with Hybrid Probabilistic Models

Abstract

Learning new tasks continuously without forgetting on a constantly changing data distribution is essential for real-world problems but extremely challenging for modern deep learning. In this work we propose HCL, a Hybrid generative-discriminative approach to Continual Learning for classification. We model the distribution of each task and each class with a normalizing flow. The flow is used to learn the data distribution, perform classification, identify task changes, and avoid forgetting, all leveraging the invertibility and exact likelihood which are uniquely enabled by the normalizing flow model. We use the generative capabilities of the flow to avoid catastrophic forgetting through generative replay and a novel functional regularization technique. For task identification, we use state-of-the-art anomaly detection techniques based on measuring the typicality of the model's statistics. We demonstrate the strong performance of HCL on a range of continual learning benchmarks such as split-MNIST, split-CIFAR, and SVHN-MNIST.

References in corpus (12)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Semi-Supervised Learning with Deep Generative Models
• Three scenarios for continual learning
• Continual Unsupervised Representation Learning
• Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting
• Task Agnostic Continual Learning via Meta Learning
• Hybrid Models with Deep and Invertible Features
• Why Normalizing Flows Fail to Detect Out-of-Distribution Data
• Gradient Projection Memory for Continual Learning
• The Effectiveness of Memory Replay in Large Scale Continual Learning
• Density of States Estimation for Out-of-Distribution Detection
• Learning latent representations across multiple data domains using Lifelong VAEGAN