Information theoretic approach to interactive learning

Abstract

The principles of statistical mechanics and information theory play an important role in learning and have inspired both theory and the design of numerous machine learning algorithms. The new aspect in this paper is a focus on integrating feedback from the learner. A quantitative approach to interactive learning and adaptive behavior is proposed, integrating model- and decision-making into one theoretical framework. This paper follows simple principles by requiring that the observer's world model and action policy should result in maximal predictive power at minimal complexity. Classes of optimal action policies and of optimal models are derived from an objective function that reflects this trade-off between prediction and complexity. The resulting optimal models then summarize, at different levels of abstraction, the process's causal organization in the presence of the learner's actions. A fundamental consequence of the proposed principle is that the learner's optimal action policies balance exploration and control as an emerging property. Interestingly, the explorative component is present in the absence of policy randomness, i.e. in the optimal deterministic behavior. This is a direct result of requiring maximal predictive power in the presence of feedback.

6 pages

References in corpus (2)

• The information bottleneck method
• Statistical Mechanics of Learning: A Variational Approach for Real Data

Cited by in corpus (24)

• Exploration in Deep Reinforcement Learning: A Survey
• Thermodynamics as a theory of decision-making with information processing costs
• The thermodynamics of prediction
• Computational Mechanics of Input-Output Processes: Structured transformations and the $\epsilon$-transducer
• The Two Kinds of Free Energy and the Bayesian Revolution
• Bounded rational decision-making from elementary computations that reduce uncertainty
• The evolution of lossy compression
• Nearly Maximally Predictive Features and Their Dimensions
• Information theory and adaptation
• Energetic cost of feedback control
• Graph-based Predictable Feature Analysis
• Learning in embodied action-perception loops through exploration
• Quantum Predictive Filtering
• Quantum correlations in predictive processes
• Novelty Search in Representational Space for Sample Efficient Exploration
• Geometry and Determinism of Optimal Stationary Control in Partially Observable Markov Decision Processes
• Partially Observable Szilard Engines
• Representing preorders with injective monotones
• Reviewing Evolution of Learning Functions and Semantic Information Measures for Understanding Deep Learning
• Circumventing the Curse of Dimensionality in Prediction: Causal Rate-Distortion for Infinite-Order Markov Processes
• Thermodynamic fluctuation theorems govern human sensorimotor learning
• Kolmogorov-Sinai entropy and dissipation in driven classical Hamiltonian systems
• Empowerment for Continuous Agent-Environment Systems
• A Non-equilibrium Thermodynamic Framework of Consciousness