Learning Human Behavior Patterns for Proactive Service System: Agglomerative Fuzzy Clustering-Based Fuzzy-State Q-learning

Modeling and recognition of human behavior patterns for proactive service system are known to be difficult. For this purpose, an agglomerative clustering-based fuzzy-state Q-learning algorithm is suggested. In the first step of the proposed method, a meaningful structure of data is discovered by using Agglomerative Iterative Bayesian Fuzzy Clustering (AIBFC). Next in the second step, the sequence of actions is learned on the basis of the structure discovered in the first step and by virtue of the proposed Fuzzy-state Q-learning (FSQL) process. These two learning steps are incorporated in an amalgamated framework of AIBFC-FSQL, which is capable of learning human behavior patterns and predicting next human actions. We show that the proposed learning method outperforms several well-known methods by conducting experiments with two real-world database.