The present disclosure relates a method and a visual interactive system for tree boosting (VISTB). Tree boosting uses a group of sequentially generated weak learners (i.e., decision trees), each learns from the mistakes of its predecessor, to push the model’s decision boundary towards the true boundary. As the number of trees keeps increasing over training, it is important to reveal how the newly added trees change the predictions of individual data instances, and how the impact of different data features evolve. To accomplish these goals, in this present disclosure, proposes a new design of temporal confusion matrix, an effective interface is provided to users to track data instances’ predictions across the tree boosting process. Also, an improved visualization is presented the users to better illustrate and compare the impact of individual data features across time. Integrating these components with a tree structure visualization component in coordinated views.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Wang, Junpeng; Zhang, Wei; Wang, Liang; and Yang, Hao, "INVESTIGATING THE EVOLUTION OF TREE BOOSTING MODELS WITH VISUAL ANALYTICS", Technical Disclosure Commons, (September 17, 2021)