XploreML: An interactive approach to study and explore machine learning models

Anonymous

Abstract:

Due to its achievements in recent years, Machine Learning (ML) is now used in a wide variety of problem domains. Educating ML has hence become an important factor to enable novel applications.
To address this challenge, this paper introduces XploreML -- an interactive approach for lecturers to teach and for students or practitioners to study and explore the fundamentals of machine learning. XploreML allows users to experiment with data preparation, data transformation and a wide range of classifiers. The data sets can be visually investigated in order to understand the complexity of the classification problem. The selected classifier can either be autonomously fitted to the training data or the effect of manually altering model hyperparameters can be explored. Additionally source code of configured ML pipelines can be extracted.
XploreML can be used within a lecture as an interactive demo or by students in a lab session. Both scenarios were evaluated with a user survey, where both variants were assessed as positive with the first yielding more positive feedback.
XploreML can be used online: ml-and-vis.shinyapps.io/XploreML

Keywords: machine learning education, visualization, classification, human-centered machine learning

TL;DR: The paper introduces an interactive approach for lecturers and students to teach, study and explore the fundamentals of machine learning.