GBDTs.jl

Grammar-Based Decision Trees is a machine learning model for the interpretable classification and categorization of data.
Author sisl
Github
Popularity
5 Stars
Updated Last
1 Year Ago
Started In
October 2017

GBDTs.jl -- Grammar-Based Decision Trees

Grammar-based decision tree (GBDT) is an interpretable machine learning model that can be used for the classification and categorization of heterogeneous multivariate time series data. GBDTs combine decision trees with a grammar framework. Each split of the decision tree is governed by a logical expression derived from a user-supplied grammar. The flexibility of the grammar framework enables GBDTs to be applied to a wide range of problems. In particular, GBDT has been previously applied to analyze multivariate heterogeneous time series data of failures in aircraft collision avoidance systems [1].

[1] Lee et al. "Interpretable Categorization of Heterogeneous Time Series Data", preprint, 2018.

Main Dependencies

  • sisl/ExprOptimization.jl
  • sisl/MultivariateTimeSeries.jl

Usage

Please see the example notebook.

Maintainers:

Build Status Coverage Status