Databricks Certified Machine Learning Associate Exam Guide

The minimally qualified candidate should be able to:

• Use Databricks Machine Learning and its capabilities within machine learning workflows, including:
  • Databricks Machine Learning (clusters, Repos, Jobs)
  • Databricks Runtime for Machine Learning (basics, libraries)
  • AutoML (classification, regression, forecasting)
  • Feature Store (basics)
  • MLflow (Tracking, Models, Model Registry)
• Implement correct decisions in machine learning workflows, including:
  • Exploratory data analysis (summary statistics, outlier removal)
  • Feature engineering (missing value imputation, one-hot-encoding)
  • Tuning (hyperparameter basics, hyperparameter parallelization)
  • Evaluation and selection (cross-validation, evaluation metrics)
• Implement machine learning solutions at scale using Spark ML and other tools, including:
  • Distributed ML Concepts
  • Spark ML Modeling APIs (data splitting, training, evaluation, estimators vs. transformers, pipelines)
  • Hyperopt
  • Pandas API on Spark
  • Pandas UDFs and Pandas Function APIs
• Understand advanced scaling characteristics of classical machine learning models, including:
  • Distributed Linear Regression
  • Distributed Decision Trees
  • Ensembling Methods (bagging, boosting)