About

This website accompanies the folliwing paper and provides an interactive way to navigate method selection across data modalities, tasks, and practical constraints.

Paper

A Practical Guide to Using Real-World Data in Artificial Intelligence Training
[Link to the paper: coming soon]

Abstract

As data grows more diverse and complex, selecting an appropriate machine learning or deep learning model for a real-world problem has become increasingly challenging. Effective model choice depends not only on predictive performance but also on the characteristics of the data and the practical constraints of the application. This paper presents a decision-oriented framework that explicitly links common data challenges to suitable classes of machine learning approaches, providing a structured perspective on model selection beyond traditional method-centric reviews. To operationalise this framework, we introduce an interactive dashboard that guides users through a structured set of questions about their data and problem setting, and maps their responses to relevant model families and supporting literature. The system translates expert knowledge into an accessible, user-facing decision-support tool, enabling more informed and transparent model selection for both practitioners and non-experts. The accompanying review synthesises key model families and recurring data challenges, forming the conceptual foundation of the proposed system. By integrating a data-centric taxonomy with an operational decision tool, this work gives researchers an accessible way to connect data characteristics to suitable modelling approaches.

About this website

This site is designed as a companion to the paper. The wizard provides a structured route through the decision framework, while the methods, concepts, and data-challenge pages provide the supporting context and literature.

Repository

The website source code is available on GitHub:
github.com/Sinamhr/data-learning-website