Lifelong Generative Modelling Using Dynamic Expansion Graph Model

TL;DR

This paper introduces DEGM, a novel lifelong learning model for VAEs that dynamically expands its architecture based on task novelty, effectively balancing performance and parameter efficiency.

Contribution

The paper presents DEGM, a dynamic expansion graph model that adapts its structure for lifelong learning, improving knowledge retention and parameter efficiency over traditional methods.

Findings

DEGM achieves optimal task performance.

DEGM minimizes the number of parameters needed.

Theoretical analysis explains VAEs' forgetting behavior.

Abstract

Variational Autoencoders (VAEs) suffer from degenerated performance, when learning several successive tasks. This is caused by catastrophic forgetting. In order to address the knowledge loss, VAEs are using either Generative Replay (GR) mechanisms or Expanding Network Architectures (ENA). In this paper we study the forgetting behaviour of VAEs using a joint GR and ENA methodology, by deriving an upper bound on the negative marginal log-likelihood. This theoretical analysis provides new insights into how VAEs forget the previously learnt knowledge during lifelong learning. The analysis indicates the best performance achieved when considering model mixtures, under the ENA framework, where there are no restrictions on the number of components. However, an ENA-based approach may require an excessive number of parameters. This motivates us to propose a novel Dynamic Expansion Graph Model (DEGM). DEGM expands its architecture, according to the novelty associated with each new databases, when compared to the information already learnt by the network from previous tasks. DEGM training optimizes knowledge structuring, characterizing the joint probabilistic representations corresponding to the past and more recently learned tasks. We demonstrate that DEGM guarantees optimal performance for each task while also minimizing the required number of parameters. Supplementary materials (SM) and source code are available in https://github.com/dtuzi123/Expansion-Graph-Model.