Supervised Learning Algorithms

It’s time to do some learning based on the data. Most people think machine learning is applying an algorithm to given data and predicting results. Well, it’s not just that simple. Eighty percent of the work involves data collection, preprocessing, cleaning, feature engineering, transformation, and selecting the best features. The remaining 20 percent is spent on building machine learning models, validation, and deployment. The entire operation is called MLOps (machine learning operations). It is similar to DevOps, but for machine learning. In order to understand and deploy a production model, you should be familiar with each component in MLOps. The previous chapters discussed the 80 percent of the work. If you skipped those chapters, we recommend reading them before you read this chapter. In addition, business and domain knowledge helps you improve the process throughout.