Latent Properties of Lifelong Learning Systems
This work addresses a methodological issue for researchers in lifelong learning AI, but it appears incremental as it focuses on improving evaluation rather than proposing a new learning paradigm.
The paper tackles the problem of confounding factors in lifelong learning metrics by introducing an algorithm-agnostic explainable surrogate-modeling approach to estimate latent properties, validated on synthetic and real data from classification and reinforcement learning tasks.
Creating artificial intelligence (AI) systems capable of demonstrating lifelong learning is a fundamental challenge, and many approaches and metrics have been proposed to analyze algorithmic properties. However, for existing lifelong learning metrics, algorithmic contributions are confounded by task and scenario structure. To mitigate this issue, we introduce an algorithm-agnostic explainable surrogate-modeling approach to estimate latent properties of lifelong learning algorithms. We validate the approach for estimating these properties via experiments on synthetic data. To validate the structure of the surrogate model, we analyze real performance data from a collection of popular lifelong learning approaches and baselines adapted for lifelong classification and lifelong reinforcement learning.