Free Practice Questions for scikit-learn Expert Practitioner Certification Certification
Study with exam-style practice questions designed to help you prepare for the scikit-learn Expert Practitioner Certification.
Start Practicing
Random Questions
Practice with randomly mixed questions from all topics
Domain Mode
Practice questions from a specific topic area
Exam Information
Exam Details
Key information about scikit-learn Expert Practitioner Certification
Exam Topics & Skills Assessed
Skills measured (from the official study guide)
Domain 1: Machine learning concepts
Subdomain 1.1: Supervised learning and unsupervised
Supervised learning and unsupervised (regression, classification, clustering, dimensional reduction)
Subdomain 1.2: Types of model families
Types of model families (tree-based, linear, ensemble, neighbors)
Subdomain 1.3: Loss functions and surrogate loss
Loss functions and surrogate loss
Subdomain 1.4: Splitting criteria in Decision Trees
Splitting criteria in Decision Trees
Subdomain 1.5: Filter, wrapper and embedded methods for feature selection
Filter, wrapper and embedded methods for feature selection
Subdomain 1.6: Calibration vs ranking power
Calibration (expected calibration error) vs ranking power (ROC AUC / GINI)
Domain 2: Model building and evaluation
Subdomain 2.1: Create your own estimator
Create your own estimator.
- NearestCentroid - Recommender systems - Transformers
Subdomain 2.2: Metadata routing
Metadata routing
Subdomain 2.3: Calibration plots and post-calibration
Calibration plots with CalibrationDisplay and post-calibration with CalibratedClassifierCV
Domain 3: Interpretation of results & communication
Subdomain 3.1: Explainability and interpretability
Explainability and interpretability.
- partial dependence plots: impact non-linear on the target? - permutation importance
Subdomain 3.2: Debugging the methodology
Debugging the methodology.
- given a plot, give a diagnostic for the model - identify pitfalls in the modeling process (e.g. Feature selection techniques inside or outside the pipeline) - code comprehension and good practices
Domain 4: Data preprocessing
Subdomain 4.1: Loading parquet datasets
Loading parquet datasets
Subdomain 4.2: Extract information from plots
Extract information from plots, e.g. decide on which family of models may be the best fit
Subdomain 4.3: Data wrangling
Data wrangling.
- Combining data from multiple sources - Adding new features or derived attributes (e.g. lagged features for time based data)
Domain 5: Model selection and validation
Subdomain 5.1: Performing hyperparameter tuning
Performing hyperparameter tuning with proper scoring rules (calibration)
Domain 6: Model deployment
Subdomain 6.1: Saving and loading trained models
Understanding how to save and load trained models using joblib , pickle or skops.
Techniques & products