"Context tree modeling of scientific data: from language to the brain" Antonio Galves

Mini-course

Antonio Galves, CEPID NeuroMat, Sao Paulo

Mini-class summary

Stochastic chains with memory of variable length, also called Variable Lenght Markov Chains and Context Tree Models.
Finding fingerprints of rhythm in writen texts as a problem of statistical model selection. A constant free algorithm for context tree model selection: the Smallest Maximizer Criterion. The case for Brazilian and European Portuguese.
Modeling neuronal networks as a system of interacting point processes with memory of variable length. How to select a graph of interactions for this class of models: a theorem and challenges.
Addressing von Helmholtz "Unconscious Inference" conjecture. Sequences of chunks of functional data driven by a context tree model. Retrieving a context tree from EEG data.

References:
A. Galves, C. Galves, J. Garcia, N. L. Garcia and F. Leonardi, Context tree selection and linguistic rhythm retrieval from written texts,. Annals Appl. Stat. vol. 6, no. 1, 186-209, 2012.
A. Galves and E. Loecherbach Infinite systems of interacting chains with memory of variable length - a stochastic model for biological neural nets, J. Stat. Phys. vol. 151, 869-921, 2013.
A. Duarte, A. Galves, G. Ost and E. Loecherbach. Estimating the interaction graph of stochastic neural dynamics, Bernoulli, to appear, 2018.
A. Duarte, R. Fraiman, A. Galves, G. Ost and C. Vargas, Context tree selection for functionnal data, preprint, 2017.

Dates

Created on February 10, 2018