NO CODE ML: Partial Automation of Data Analysis

Abstract

Machine Learning, a core facet of Artificial Intelligence and Computer Science, leverages information and algorithms to enhance learning and improve precision. The NoCode Machine Learning project presents a web-based application that empowers users to execute diverse machine learning tasks without manual coding. Its intuitive user interface caters to domain specialists, business analysts, and machine learning enthusiasts, enabling them to effortlessly upload datasets, preprocess data, train models, and evaluate performance. This paper seeks to conduct an in-depth analysis of the potential impact of the NoCode Machine Learning project on the Data Science Industry. It examines the application's features, usability, and performance, while comparing its efficacy against conventional code-based methodologies. The NoCode ML project stands as a pioneering endeavor, as it forges a code-free pathway to interact with ML algorithms through a graphical interface. The paper explores the pivotal steps in the machine learning process, encompassing dataset uploading, data preprocessing, model training, and evaluation, utilizing metrics like Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). Additionally, the NoCode ML project boasts a Tabler Dashboard, equipped with project management tools to track progress, milestones, and tasks. While the application demonstrates numerous advantages, it also faces certain challenges, such as limited model options and the necessity of external data preprocessing. In conclusion, the NoCode Machine Learning project emerges as a valuable tool, revolutionizing machine learning practices by rendering them accessible and dispelling the barriers of coding. It offers a user-friendly platform, propelling diverse users to wield the power of machine learning effectively and contribute to the ever-evolving landscape of knowledge and technology.