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.

On Multi-dimensional Markov Chain Models

A New Multivariate Markov Chain Model for Adding a New Categorical Data Sequence

A Simplified Higher-Order Markov Chain Model

On Computation With Higher-Order Markov Chains

GenMarkov: Modeling Generalized Multivariate Markov Chains in R

Modeling User Navigation Sequences Based on Multi-Markov Chains

Markov Chains and Monte Carlo Simulation

An improved multivariate Markov chain model for credit risk

A Parsimonious Multivariate Markov Chain Model for Credit Risk

Max Markov Chain

Dynamical mixture modeling with fast, automatic determination of Markov chains

A new belief Markov chain model and its application in inventory prediction

Learning Mixtures of Markov Chains from Aggregate Data with Structural Constraints

A tutorial on pharmacometric Markov models

Markov Chain Combination Prediction Model and Its Application in Stock Market

Modeling Regular Occupancy in Commercial Buildings Using Stochastic Models

Markov chain random fields, spatial Bayesian networks, and optimal neighborhoods for simulation of categorical fields

Low-Rank Tensors for Multi-Dimensional Markov Models

Stochastic Relational Models for Large-scale Dyadic Data using MCMC

Multi-label Collective Classification Via Markov Chain Based Learning Method

A Coupled Markov Chain Approach to Credit Risk Modeling