Databricks ML Assoc study guide · 2026

The 4 Databricks ML Assoc domains, explained

4 domains and 48 official objectives. Databricks Machine Learning carries the most weight at 38%. Each guide explains what the domain covers, how its questions are framed and where candidates lose marks.

Start with Domain 1 See all domains
Questions
48 scored (up to 10 additional unscored items)
Time
90 minutes
Passing score
Not published by Databricks
Outline
Machine Learning Associate exam guide, file labelled August 2026 (content dated 1 March 2025) (current)

Exam blueprint

Bar widths follow the official weights. Weight ranges are drawn at their midpoint.

  1. 1 38%
  2. 2 19%
  3. 3 31%
  4. 4 12%
  1. 1 Databricks Machine Learning
  2. 2 Data Processing
  3. 3 Model Development
  4. 4 Model Deployment

All 4 domains

48 official objectives 38–52 hrs suggested study

  1. Domain 01 38%

    Databricks Machine Learning

    The largest section of the exam at 38%. It covers MLOps practice, Databricks Runtime for Machine Learning, AutoML, feature tables in Unity Catalog, MLflow experiment tracking, and registering, tagging and promoting models with aliases in the Unity Catalog registry.

    Inside this domain

    • Identify the best practices of an MLOps strategy
    • Identify the advantages of using ML runtimes
    • Identify how AutoML facilitates model/feature selection.
    • Identify the advantages AutoML brings to the model development process
    • and 14 more objectives

    Watch for Sort RMSE descending

    Objectives
    18
    Questions est.
    ~18
    Study time est.
    14–18 hrs
  2. Domain 02 19%

    Data Processing

    This section, worth 19%, tests how you explore and prepare data in Spark before modelling: summary statistics, outlier removal, plots and feature comparisons, choosing and applying imputation, one-hot encoding, and when a log transform helps.

    Inside this domain

    • Compute summary statistics on a Spark DataFrame using .summary() or dbutils data summaries
    • Remove outliers from a Spark DataFrame based on standard deviation or IQR
    • Create visualizations for categorical or continuous features
    • Compare two categorical or two continuous features using the appropriate method
    • and 5 more objectives

    Watch for Mean for skewed data

    Objectives
    9
    Questions est.
    ~9
    Study time est.
    7–10 hrs
  3. Domain 03 31%

    Model Development

    The second-largest section at 31%. It tests algorithm selection, handling class imbalance, Spark ML estimators, transformers and pipelines, hyperparameter tuning with Hyperopt and search strategies, cross-validation, evaluation metrics, and the bias-variance tradeoff.

    Inside this domain

    • Use ML foundations to select the appropriate algorithm for a given model scenario
    • Identify methods to mitigate data imbalance in training data
    • Compare estimators and transformers
    • Develop a training pipeline
    • and 11 more objectives

    Watch for Forget the refit

    Objectives
    15
    Questions est.
    ~15
    Study time est.
    12–16 hrs
  4. Domain 04 12%

    Model Deployment

    The smallest section at 12%. It covers choosing batch, streaming or real-time serving, deploying a custom MLflow model to a serving endpoint, batch inference with pandas, streaming inference in Delta Live Tables pipelines, querying endpoints and splitting traffic.

    Inside this domain

    • Identify the differences and advantages of model serving approaches: batch, realtime, and streaming
    • Deploy a custom model to a model endpoint
    • Use pandas to perform batch inference
    • Identify how streaming inference is performed with Delta Live Tables
    • and 2 more objectives

    Watch for Endpoint for nightly scoring

    Objectives
    6
    Questions est.
    ~6
    Study time est.
    5–8 hrs

Study the heaviest domains first

Marks follow weight. This order puts the most marks behind you first, across about 38–52 hrs of study.

  1. 1 Databricks Machine Learning 38% · 14–18 hrs
  2. 3 Model Development 31% · 12–16 hrs
  3. 2 Data Processing 19% · 7–10 hrs
  4. 4 Model Deployment 12% · 5–8 hrs

Try one

Practice questionDomain 1: Databricks Machine Learning

A data scientist registered version 4 of main.fraud.classifier in Unity Catalog. A nightly batch job loads models:/main.fraud.classifier@champion, which currently points to version 3. After validation, the job should use version 4 without any change to its code. What should the data scientist do?

One trap from every domain

The answer that looks right, and the clue that gives it away.

  • Domain 1

    Sort RMSE descending

    Clue The metric is RMSE, MAE or log loss.

  • Domain 2

    Mean for skewed data

    Clue The column is described as skewed or containing extreme values.

  • Domain 3

    Forget the refit

    Clue The question says the best model is produced or asks for all models trained.

  • Domain 4

    Endpoint for nightly scoring

    Clue Results are consumed later, such as next morning.

Other Artificial Intelligence & Machine Learning certifications

Study guides for exams that overlap with Databricks ML Assoc.

Practise what you read

Databricks ML Assoc domains: common questions

How many domains are in the Databricks ML Assoc exam?

Databricks ML Assoc has 4 domains with 48 official objectives between them: Databricks Machine Learning, Data Processing, Model Development and Model Deployment.

Which Databricks ML Assoc domain carries the most weight?

Databricks Machine Learning is the heaviest at 38%, followed by Model Development (31%). The lightest is Model Deployment (12%). Weights come from Databricks's official outline.

What is the Databricks ML Assoc exam format?

Questions: 48 scored (up to 10 additional unscored items). Time allowed: 90 minutes. Passing score: Not published by Databricks.

How long should I study for Databricks ML Assoc?

Our planning estimate is 38–52 hrs across all domains, split by weight. It is an editorial estimate, not an official figure, so adjust it to your experience and to how you score on practice questions.

Where should I start?

Start with Databricks Machine Learning, because it carries the most marks, then work down by weight. Use the checklist on each domain page to confirm you can do what the domain asks before moving on.

Which outline is this guide based on?

This guide follows Machine Learning Associate exam guide, file labelled August 2026 (content dated 1 March 2025) (current), published by Databricks. It was last reviewed in October 2026.

Domain names, weights and objectives come from Databricks's official outline (Machine Learning Associate exam guide, file labelled August 2026 (content dated 1 March 2025) (current)). Question counts and study hours are editorial planning estimates, not official figures. Last reviewed October 2026.