[NEW] Databricks Certified Machine Learning Associate

ML Workflows (19%) Building pipelines with Databricks Jobs, Orchestrating ML tasks with Delta Lake, Monitoring ML pipeline

Model Development (31%) Preparing data for training, Training models, Hyper-parameter tuning, Model evaluation, Model selection

Model Deployment (12%) Register and serve models, Deploy models using MLflow Model Registry, Scaling model serving, Monitoring model performance

B. Databricks Feature Store

C. MLflow Tracking

D. Databricks AutoML

E. Delta Live Tables

F. MLflow Model Registry

Correct Answer: C. MLflow Tracking

Explanation:

B. Databricks Feature Store is incorrect because it is used to discover and share features across machine learning models.

C. MLflow Tracking is correct because it provides an API and UI specifically built for logging parameters, code versions, metrics, and output files when running machine learning code.

D. Databricks AutoML is incorrect because it automates the process of training models rather than serving as the underlying tracking log.

E. Delta Live Tables is incorrect because it is a framework for building reliable data pipelines, not for tracking ML experiments.

F. MLflow Model Registry is incorrect because it is a centralized model store used to manage the lifecycle of models after they are trained, not the tracking component used during the training phase.

B. Python notebooks containing the source code for each trial run

C. A direct streaming connection to external Kafka topics

D. Translated SQL queries into PySpark DataFrames

E. A centralized data warehouse architecture

F. Manual REST API deployment scripts for external servers

Correct Answer: B. Python notebooks containing the source code for each trial run

Explanation:

B. Python notebooks containing the source code for each trial run is correct because Databricks AutoML adopts a "glass box" approach, generating editable notebooks for each model trial so data scientists can review and tweak the code.

C. A direct streaming connection to external Kafka topics is incorrect because AutoML operates on static datasets for training, not direct streaming pipelines.

D. Translated SQL queries into PySpark DataFrames is incorrect as this relates to query execution and compilation, not the machine learning automation process.

E. A centralized data warehouse architecture is incorrect because AutoML builds models, not data storage architectures.

F. Manual REST API deployment scripts is incorrect because Databricks handles deployment through MLflow Model Serving rather than generating manual external server scripts.

B. Feature Store

C. MLflow Model Registry

D. Delta Lake Time Travel

E. MLflow Tracking

F. Databricks SQL Warehouses

Correct Answer: C. MLflow Model Registry

Explanation:

B. Feature Store is incorrect because it manages machine learning features for training and inference, not the deployment status of the models themselves.

C. MLflow Model Registry is correct because it is a centralized repository that provides model lineage, versioning, and stage transitions from Staging to Production.

D. Delta Lake Time Travel is incorrect because it allows you to query older versions of data tables, not manage machine learning model stages.

E. MLflow Tracking is incorrect because it logs experiments and metrics during the development phase, whereas the Registry handles the deployment stages.

F. Databricks SQL Warehouses is incorrect because they are compute resources for running SQL queries, entirely unrelated to model versioning.

You can retake the exams as many times as you want

This is a huge original question bank

You get support from me if you have questions

Each question has a detailed explanation

Mobile-compatible with the Udemy app