Discovering Interpretable Machine Learning Models in Parallel Coordinates
This provides a visual knowledge discovery method for end-users in interpretable machine learning, though it appears incremental as it builds on existing concepts like hypercubes and decision trees.
The paper tackles the problem of interpretable machine learning by proposing the Hyper algorithm, which discovers hyper-blocks in parallel coordinates for classification tasks, showing it generalizes decision trees and avoids overgeneralization and overfitting.
This paper contributes to interpretable machine learning via visual knowledge discovery in parallel coordinates. The concepts of hypercubes and hyper-blocks are used as easily understandable by end-users in the visual form in parallel coordinates. The Hyper algorithm for classification with mixed and pure hyper-blocks (HBs) is proposed to discover hyper-blocks interactively and automatically in individual, multiple, overlapping, and non-overlapping setting. The combination of hyper-blocks with linguistic description of visual patterns is presented too. It is shown that Hyper models generalize decision trees. The Hyper algorithm was tested on the benchmark data from UCI ML repository. It allowed discovering pure and mixed HBs with all data and then with 10-fold cross validation. The links between hyper-blocks, dimension reduction and visualization are established. Major benefits of hyper-block technology and the Hyper algorithm are in their ability to discover and observe hyper-blocks by end-users including side by side visualizations making patterns visible for all classes. Another advantage of sets of HBs relative to the decision trees is the ability to avoid both data overgeneralization and overfitting.