Databricks - Machine Learning Associate Study Guide

2: ML Workflows

This chapter covers the data-processing workflow that precedes model training on Databricks: profiling a Spark DataFrame with summary statistics, detecting and removing outliers with the standard-deviation and interquartile-range methods, visualizing categorical and continuous features, testing relationships between feature pairs, imputing missing values, applying one-hot encoding, and deciding when a log-scale transformation helps. Every technique is shown in Python with PySpark, scikit-learn, and pandas so the code matches what the exam presents.