Pycaret is an automated machine learning library that automates the complete pipeline of the machine learning life cycle by implementing all the phases from preprocessing of data to the model deployment.
You need to simply pass the data and the library will take care of all the preprocessing steps like handling missing values, removing outliers, fixing class imbalance, scaling the features, encoding, splitting, etc.
Then model building with cross-validation and selecting the best one based on the evaluation metrics. Further, you can also tune the model for improving a specified metric as well and deploy it on AWS optionally.
Let’s get started by installing the library
Importing the required libraries and loading a built-in dataset.
Setting up the pipelines and comparing the models based on their results and visualizing the classification report.
Tuning the hyperparameter for improving the results and saving it.