Making Sense of Deployment: Feature Engineering, Training, Testing, and Monitoring

Congratulations! You’ve spent a considerable amount of time working on getting your data just right and now you’re ready to leave that step behind and focus on your machine learning models. Well, yes and no. Now it’s time to actually start mining your unified dataset for insights and using it to build models that will give you the best results to your predictive questions. Part three of our series Explorium Explains: Data for Machine Learning, shines a light on deployment. In this whitepaper, we’ll dive deep into manual vs. automated feature engineering techniques, get technical with an in-depth look at the best ways to split your data for training and different testing methodologies, and cover how to put a plan in place to ensure your models stay accurate over time.