Automated Machine Learning-Algorithm Selection with Fine-Tuned Parameters

In recent years, the impact and usage of Machine Learning have increased multi-fold. To get the best outcome, it is imperative to choose the right algorithm. Hence, Algorithm Selection is one of the most important stages in the Machine Learning Lifecycle. Selecting the optimal algorithm out of many possible algorithms for a given problem statement is always a tedious and time-consuming process. This project assists the researchers and Machine Learning enthusiasts in selecting the appropriate algorithm by providing a comparative selection among the different machine learning algorithms available for a pre-processed dataset. Besides the best Algorithm Selection, this project also aims to provide the best hyper-parameters for the opted algorithm using a novel technique. The Algorithm Selection functionality is furnished by comparing the various accuracy metrics like Accuracy, Recall, Precision, F1 Score, R2 Score, Mean Squared Error, etc. The objective of this project is to compare numerous algorithms and suggest the best algorithm. Thus, the user can instead concentrate on Data Collection and Data Pre-processing to achieve success in building the best model.