Variable Length Markov Chains: Methodology, Computing, and Software

By Martin Mächler, Peter Buhlmann. 2004

378 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/80a9e398710b7d2346f697c0b68206a698c12d9f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Cosma Rohilla Shalizi
2. Hendrik Purwins
3. Kristina Lisa Shalizi
4. B. Kedem
5. Ricard Marxer
6. Konstantinos Fokianos
7. Miguel de Carvalho
8. A. C. Davison

In Molecular Biology

1. Hendrik Purwins
2. Cosma Rohilla Shalizi
3. Kristina Lisa Shalizi
4. Ricard Marxer
5. Konstantinos Fokianos

Citations in the corpus

• Blind Construction of Optimal Nonlinear Recursive Predictors for Discrete Sequences (2004, methodology): “The “context” algorithm of Rissanen [18] and its descendants [19, 20, 21, 22] construct “variable-length Markov models” (VLMMs) from sequence data.”
• Regression Theory for Categorical Time Series (2003, background): “Model (26) is less parsimonious than model (25) in the sense that it requires m(m − 1) + 1 additional parameters for each lag. However, it accommodates a dynamic relation between each lag and time period. The work in Le, Martin and Raftery (1996) and, more recently, in Wong and Li (2000) extends def”
• Unsupervised Incremental Online Learning and Prediction of Musical Audio Signals (2015, methodology): “As alternatives to CB and HN, variable length Markov models [47], [13], [46] or other deep learning architectures can be used, thereby overcoming the context length limitation of HN and CB and the slow learning of CB.”