Databricks Machine Learning Associate

Using the Databricks Machine Learning platform: clusters and ML runtime, AutoML, the Feature Store, MLflow tracking, and the Model Registry.

• Databricks ML runtime and ML-optimized clusters
• AutoML for baseline models and generated notebooks
• Databricks Feature Store / Feature Engineering basics
• MLflow Tracking: experiments, runs, autologging
• MLflow Model Registry basics (registering and versioning models)
• Repos, notebooks, and jobs for ML workflows

Free Databricks Machine Learning practice questions →Databricks Machine Learning study guide →

Core ML workflow steps on Databricks: exploratory data analysis, feature engineering, data splitting, and preparing data for training.

• Exploratory data analysis (summary statistics, distributions)
• Handling missing values and outliers
• Feature engineering: encoding, scaling, and transformations
• Train/validation/test splitting and cross-validation
• Addressing class imbalance

Free ML Workflows practice questions →ML Workflows study guide →

Training and tuning models on Databricks: Spark ML and single-node libraries, hyperparameter tuning with Hyperopt, and evaluation metrics.

• Training models with scikit-learn/Spark ML on Databricks
• Spark ML Pipelines (Transformers, Estimators)
• Hyperparameter tuning with Hyperopt (and SparkTrials)
• Evaluation metrics (accuracy, precision/recall, F1, RMSE, AUC)
• Overfitting/underfitting, regularization, and validation
• Tracking training with MLflow

Free Model Development practice questions →Model Development study guide →

Deploying models on Databricks: batch, streaming, and real-time serving paradigms and using the Model Registry for deployment.

• Batch inference (applying a model to a dataset)
• Streaming inference with Structured Streaming
• Real-time serving with Databricks Model Serving
• Loading registered models for inference
• Choosing a deployment approach by latency and volume

Free Model Deployment practice questions →Model Deployment study guide →