Abstract:Statistical models based on Markov chains, especially mixtures of Markov chains, have recently been studied and demonstrated to be effective in various data mining applications such as tourist flow analysis, animal migration modeling, and transportation administration. Nevertheless, the research so far has mainly focused on analyzing data at individual levels. Due to security and privacy reasons, however, the observations in practice usually consist of coarse-grained statistics of individual data, a.k.a. aggregate data, rendering learning mixtures of Markov chains an even more challenging problem. In this work, we show that this challenging problem, although intractable in its original form, can be solved approximately by posing structural constraints on the transition matrices. The proposed structural constraints include specifying active state sets corresponding to the chains and adding a pairwise sparse regularization term on transition matrices. Based on these two structural constraints, we propose a constrained least-squares method to learn mixtures of Markov chains. We further develop a novel iterative algorithm that decomposes the overall problem into a set of convex subproblems and solves each subproblem efficiently, making it possible to effectively learn mixtures of Markov chains from aggregate data. We propose a framework for generating synthetic data and analyze the complexity of our algorithm. Additionally, the empirical results of the convergence and the robustness of our algorithm are also presented. These results demonstrate the effectiveness and efficiency of the proposed algorithm, comparing with traditional methods. Experimental results on real-world data sets further validate that our algorithm can be used to solve practical problems.

Markovletics: Methods and A Novel Application for Learning Continuous-Time Markov Chain Mixtures

ULTRA-MC: A Unified Approach to Learning Mixtures of Markov Chains via Hitting Times

Learning Mixtures of Markov Chains with Quality Guarantees

Dynamical mixture modeling with fast, automatic determination of Markov chains

Learning Mixtures of Markov Chains from Aggregate Data with Structural Constraints

Laplace Transform Based Low-Complexity Learning of Continuous Markov Semigroups

Learning Markovian dynamics with spectral maps

Evolving Markov Chains: Unsupervised Mode Discovery and Recognition from Data Streams

A Tutorial on the Spectral Theory of Markov Chains

Adapting to Mixing Time in Stochastic Optimization with Markovian Data

Neural Markov Jump Processes

Markov Chain Monte Carlo for Continuous-Time Switching Dynamical Systems

Learning an evolved mixture model for task-free continual learning

Bounds on Lifting Continuous Markov Chains to Speed Up Mixing

An introduction to the Markov chain Monte Carlo method

MCMC-driven learning

Empirical and Instance-Dependent Estimation of Markov Chain and Mixing Time

Updating Markov chain models using the ensemble Kalman filter

Interacting Markov chain Monte Carlo methods for solving nonlinear measure-valued equations

Markov Chain Monte Carlo in Practice