Behavior Learning (BL): Learning Hierarchical Optimization Structures from Data
2026-02-23 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors introduce Behavior Learning (BL), a new machine learning method that models data using building blocks based on decision-making principles from behavioral science. Each block can be understood as solving an optimization problem, making the overall model both understandable and identifiable. Their approach works on simple or complex, layered problems and can predict and generate data effectively. They prove mathematically that BL can approximate a wide range of functions and that their identifiable version (IBL) has good statistical properties. In tests, BL showed good prediction accuracy, interpretability, and worked well with high-dimensional data.
Behavior Learning (BL)Utility Maximization Problem (UMP)InterpretabilityIdentifiabilityHierarchical OptimizationUniversal ApproximationM-estimationPredictive ModelingCompositional ModelsHigh-dimensional Data
Authors
Zhenyao Ma, Yue Liang, Dongxu Li
Abstract
Inspired by behavioral science, we propose Behavior Learning (BL), a novel general-purpose machine learning framework that learns interpretable and identifiable optimization structures from data, ranging from single optimization problems to hierarchical compositions. It unifies predictive performance, intrinsic interpretability, and identifiability, with broad applicability to scientific domains involving optimization. BL parameterizes a compositional utility function built from intrinsically interpretable modular blocks, which induces a data distribution for prediction and generation. Each block represents and can be written in symbolic form as a utility maximization problem (UMP), a foundational paradigm in behavioral science and a universal framework of optimization. BL supports architectures ranging from a single UMP to hierarchical compositions, the latter modeling hierarchical optimization structures. Its smooth and monotone variant (IBL) guarantees identifiability. Theoretically, we establish the universal approximation property of BL, and analyze the M-estimation properties of IBL. Empirically, BL demonstrates strong predictive performance, intrinsic interpretability and scalability to high-dimensional data. Code: https://github.com/MoonYLiang/Behavior-Learning ; install via pip install blnetwork.