Web Based Machine Learning Automated Pipeline

Abstract

Abstract: With the increasing volume, velocity, veracity, and variety of data, it has become critical to have efficient techniques and tools for managing and analyzing data in machine learning. Abstraction is a powerful concept that allows users to interact with machine learning algorithms without understanding their technical implementation details. In this project the user will provide the dataset in .csv format the dataset is then processed further to different machine learning preprocessing steps like removing unwanted columns, handling missing values, label encoding, outlier detection and removal, normalization, model building, model prediction, and the result can be downloaded as pdf, tracable pdf and CSV, this all processes gives a result of different model and their respective accuracy so that we can choose the best model for that particular dataset. tracable pdf will be containing all the timestamp of the processes done with their respective result, Apart from client-server model user is also provided a api so that all processes can be implemented in different platforms like c++, java, ruby etc. Overall, this paper highlights the critical role of abstraction in managing the complexity of data and machine learning algorithms, enabling more efficient and effective analysis of large and complex datasets.