Markov-switching decision trees
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Standard
Markov-switching decision trees. / Adam, Timo; Ötting, Marius; Michels, Rouven.
I: AStA Advances in Statistical Analysis, 2024.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Markov-switching decision trees
AU - Adam, Timo
AU - Ötting, Marius
AU - Michels, Rouven
N1 - Publisher Copyright: © The Author(s) 2024.
PY - 2024
Y1 - 2024
N2 - Decision trees constitute a simple yet powerful and interpretable machine learning tool. While tree-based methods are designed only for cross-sectional data, we propose an approach that combines decision trees with time series modeling and thereby bridges the gap between machine learning and statistics. In particular, we combine decision trees with hidden Markov models where, for any time point, an underlying (hidden) Markov chain selects the tree that generates the corresponding observation. We propose an estimation approach that is based on the expectation-maximisation algorithm and assess its feasibility in simulation experiments. In our real-data application, we use eight seasons of National Football League (NFL) data to predict play calls conditional on covariates, such as the current quarter and the score, where the model’s states can be linked to the teams’ strategies. R code that implements the proposed method is available on GitHub.
AB - Decision trees constitute a simple yet powerful and interpretable machine learning tool. While tree-based methods are designed only for cross-sectional data, we propose an approach that combines decision trees with time series modeling and thereby bridges the gap between machine learning and statistics. In particular, we combine decision trees with hidden Markov models where, for any time point, an underlying (hidden) Markov chain selects the tree that generates the corresponding observation. We propose an estimation approach that is based on the expectation-maximisation algorithm and assess its feasibility in simulation experiments. In our real-data application, we use eight seasons of National Football League (NFL) data to predict play calls conditional on covariates, such as the current quarter and the score, where the model’s states can be linked to the teams’ strategies. R code that implements the proposed method is available on GitHub.
KW - Decision trees
KW - EM algorithm
KW - Hidden Markov models
KW - Time series modeling
UR - http://www.scopus.com/inward/record.url?scp=85194720160&partnerID=8YFLogxK
U2 - 10.1007/s10182-024-00501-6
DO - 10.1007/s10182-024-00501-6
M3 - Journal article
AN - SCOPUS:85194720160
JO - AStA Advances in Statistical Analysis
JF - AStA Advances in Statistical Analysis
SN - 1863-8171
ER -
ID: 395079319