Facilitating Machine Learning Model Comparison and Explanation through a Radial Visualisation

Building an effective Machine Learning (ML) model for a data set is difficult task involving various steps. One of the most important steps to compare substantial amount generated ML models find optimal one deployment. It challenging such with dynamic number features. Comparison more than only finding differences performance, as users are also interested in relations between features and performance feature importance explanations. This paper proposes RadialNet Chart, novel visualisation approach, trained different given while revealing implicit dependent relations. In represented by lines arcs, respectively. These effectively using recursive function. The dependence encoded into structure visualisation, where their directly revealed from related line connections. information colour width Chart. Taken together can be discerned Chart Compared other commonly used approaches, help simplify comparison process benefits following: efficient terms helping focus attention visual elements interest easier discern visually instead through complex algorithmic calculations