Development of an AutoML web application for machine learning tasks

Automated Machine Learning, also known as AutoML, is a new and florid field of data science that aims at symplifying the process of designing, development and usage of typical machine learning models. Usually the tasks involved when model-building are performed by experts in the field of computer science, expecially due to the technicalities such as, but not only: data preprocessing, feature engineering, model selection, hyperparameter tuning, model evaluation, performance analysis and forecasting procedures. These are literal barriers that limit the accessibility for non-expert users. Exploiting AutoML systems aim to automate these processes, making machine learning available to all users, regarding of their knowledge. This thesis navigate the role of AutoML in ”democratizing” machine learning, emphasizing on simplicity, usability and efficiency. It examines challenges usually associated with automating end-to-end machine learning pipelines. Additionally, this study highlights the integration of a chatbot assistant powered by a large language model (LLM) that guides the user in its journey as a machine learning expert. The research culminates in the development of a web application based on AutoML, designed to implement machine learning tasks for users with varying levels of expertise. This work is a contribution to the broader adoption of machine learning across diverse industries, as the applications are fair and diverse, regarding of the field of application.