Mixture Markov Model

Mixture Markov models are statistical tools used to analyze data exhibiting hidden subgroups and sequential dependencies, aiming to identify these subgroups and model their dynamic behavior. Current research focuses on improving algorithm efficiency, particularly addressing challenges like local optima in Expectation-Maximization (EM) algorithms and developing robust methods for high-dimensional data with minimal separation between components. These models find applications in diverse fields, including student engagement analysis, where they reveal patterns in learning behaviors, and more broadly in clustering problems across various scientific domains. The development of efficient and robust algorithms is crucial for expanding the applicability of mixture Markov models to increasingly complex datasets.

Papers

October 18, 2024

Provable In-context Learning for Mixture of Linear Regressions using Transformers
Yanhao Jin, Krishnakumar Balasubramanian, Lifeng Lai
Transformer Megatron Decepticons Context Learning Mixture Component Attention Layer Expectation Maximization Linear Regression Risk Bound Mixture Markov Model

February 10, 2024

Modeling and predicting students' engagement behaviors using mixture Markov models
R. Maqsood, P. Ceravolo, C. Romero, S. Ventura
Expectation Maximization Student Friendly Knowledge Student Engagement Model Based Clustering Engagement Behavior E Learning Mixture Markov Model

December 19, 2023

Clustering Mixtures of Bounded Covariance Distributions Under Optimal Separation
Ilias Diakonikolas, Daniel M. Kane, Jasper C. H. Lee, Thanasis Pittas
Fine Grained Separation Performance Optimal Separation Mixture Weight Identity Covariance Mixture Markov Model

January 11, 2023

Private estimation algorithms for stochastic block models and mixture models
Hongjie Chen, Vincent Cohen-Addad, Tommaso d'Orsi, Alessandro Epasto, Jacob Imola, David Steurer, Stefan Tiegel
Mixture Model Private Algorithm Stochastic Block Model Private Estimation Spherical Gaussian Mixture Markov Model

October 25, 2022

Analysis of Estimating the Bayes Rule for Gaussian Mixture Models with a Specified Missing-Data Mechanism
Ziyang Lyu
General Analysis Semi Supervised Learning Estimation Task Gaussian Mixture Model Bayes Rule Class Conditional Missingness Shift Mixture Markov Model Missing Data Mechanism

July 26, 2022

Efficient Algorithms for Sparse Moment Problems without Separation
Zhiyuan Fan, Jian Li
Gaussian Mixture Separation Performance Efficient Algorithm Moment Problem Mixture Markov Model

July 22, 2022

Generalized Identifiability Bounds for Mixture Models with Grouped Samples
Robert A. Vandermeulen, René Saitenmacher
Mixture Model Novel Identifiability Test Partial Identifiability Finite Mixture Independent Sampling Mixture Markov Model Multinomial Mixture

Mixture Markov Model

Papers

Provable In-context Learning for Mixture of Linear Regressions using Transformers

Modeling and predicting students' engagement behaviors using mixture Markov models

Clustering Mixtures of Bounded Covariance Distributions Under Optimal Separation

Private estimation algorithms for stochastic block models and mixture models

Analysis of Estimating the Bayes Rule for Gaussian Mixture Models with a Specified Missing-Data Mechanism

Efficient Algorithms for Sparse Moment Problems without Separation

Generalized Identifiability Bounds for Mixture Models with Grouped Samples