ML Development

ML Development Pipeline In addition to training, the process of developing a new ML Model to solve a problem has a lot of steps if you want to do it right. See GeeksForGeeks Machine Learning Tutorial for links to how-tos on each step.

Data Preprocessing

Getting and cleaning the data, figuring out which features to use, potentially combine them to reduce the number of features while retaining the important information. Includes scaling and transforming.

Exploratory Data Analysis

This seems like it should have a lot of overlap with the preprocessing step since it’s about looking at features individually or looking at relationships between multiple features. Seems related to feature engineering and feature extraction, but

Model Evaluation and Tuning

Look at how well the model performed, tune its parameters improve its performance.

Deploying

Once a Machine Learning model is trained with satisfactory performance, it needs to be deployed somewhere to be useful. You might deploy it with a Streamlit App or set up an API with Flask or FastAPI so other applications can access it.