Periodic split method: learning more readable decision trees for human activities
This paper presents specific features of the craft ai Machine Learning engine that enable it to better take into account typical rhythms in human activities. In particular, it improves the quality and explainability of predictive models related to those.
This work was presented at APIA 2017 in Caen (France) and published in its proceedings.
Placing your trust in algorithms is a major issue in society today. This article introduces a novel split method for decision tree generation algorithms aimed at improving the quality/readability ratio of generated decision trees. We focus on human activities learning that allow the definition of new temporal features. By virtue of these features, we present here the periodic split method, which produces similar or superior quality trees with reduced tree depth.
Read the full paper HERE