CDS
Coming Soon
Expected availability announced soon
This course is in active development. Preview the scope below and create a free account to be notified the moment it goes live.
CDS
The ICCP Certified Data Scientist (CDS) exam validates professionals' ability to apply data science methodology, statistical foundations, machine learning, deep learning, and MLOps to deliver production‑ready solutions.
90
Minutes
Who Should Take This
Mid‑career data analysts, machine‑learning engineers, and analytics consultants with two to ten years of hands‑on experience benefit most. They seek formal recognition of their expertise, aim to deepen statistical and advanced modeling skills, and need proven practices for deploying models at scale in enterprise environments.
What's Covered
1
Domain 1: Data Science Methodology and Process
2
Domain 2: Statistical Foundations
3
Domain 3: Machine Learning
4
Domain 4: Deep Learning and Advanced Analytics
5
Domain 5: Model Deployment and MLOps
6
Domain 6: Ethics, Governance, and Communication
What's Included in AccelaStudy® AI
Adaptive Knowledge Graph
Practice Questions
Lesson Modules
Console Simulator Labs
Exam Tips & Strategy
20 Activity Formats
Course Outline
56 learning goals
1
Domain 1: Data Science Methodology and Process
3 topics
Data science lifecycle and problem formulation
- Apply the CRISP-DM methodology to structure a data science project through business understanding, data understanding, data preparation, modeling, evaluation, and deployment phases.
- Analyze a business problem to determine whether it is amenable to data science solutions, defining success metrics, feasibility constraints, and expected business impact.
- Design an experimental framework for data science projects that defines hypotheses, control variables, evaluation criteria, and statistical significance thresholds.
Data exploration and preparation
- Apply exploratory data analysis techniques including summary statistics, distribution analysis, correlation matrices, and visualization to understand data characteristics and identify patterns.
- Implement data preparation pipelines including missing value imputation, outlier treatment, feature scaling, encoding categorical variables, and handling imbalanced datasets.
- Analyze data quality and completeness to assess fitness for modeling, identifying sampling biases, measurement errors, and data leakage risks that could invalidate results.
Data science project management
- Apply Agile data science project management including sprint planning for experimentation, backlog management for model iterations, and stakeholder review cadences.
- Analyze data science project risks including data availability delays, model performance plateaus, scope creep, and production deployment challenges to develop mitigation plans.
- Design a data science project governance framework that establishes review gates, model approval processes, and success criteria for advancing from experimentation to production deployment.
2
Domain 2: Statistical Foundations
3 topics
Probability and statistical inference
- Apply probability theory including conditional probability, Bayes theorem, probability distributions, and the central limit theorem to data science modeling scenarios.
- Implement hypothesis testing procedures including t-tests, chi-square tests, ANOVA, and non-parametric alternatives with appropriate significance levels and power analysis.
- Analyze the validity of statistical conclusions by evaluating assumptions, sample size adequacy, multiple comparison corrections, and practical versus statistical significance.
Regression and predictive modeling
- Implement linear and logistic regression models including coefficient interpretation, multicollinearity diagnosis, regularization techniques, and model diagnostics.
- Apply time series analysis techniques including stationarity testing, ARIMA modeling, seasonal decomposition, and forecasting with confidence intervals.
- Analyze regression model performance using residual diagnostics, cross-validation, information criteria, and predictive accuracy metrics to select the best model specification.
- Design a Bayesian inference approach for parameter estimation and prediction, selecting appropriate priors and evaluating posterior distributions for model-based decisions.
Experimental design and causal inference
- Implement A/B testing frameworks including randomization, sample size calculation, duration estimation, and statistical analysis for online experiments.
- Apply causal inference techniques including difference-in-differences, instrumental variables, and propensity score matching for observational data analysis.
- Analyze experimental results by evaluating effect sizes, confidence intervals, p-values, and practical significance to make data-driven business recommendations.
3
Domain 3: Machine Learning
3 topics
Supervised learning algorithms
- Implement classification algorithms including decision trees, random forests, gradient boosting, support vector machines, and k-nearest neighbors with appropriate hyperparameter tuning.
- Apply feature engineering techniques including feature selection, polynomial features, interaction terms, binning, and domain-specific transformations to improve model performance.
- Analyze classification model performance using confusion matrices, ROC-AUC curves, precision-recall trade-offs, and calibration plots to select optimal decision thresholds.
- Design an ensemble learning strategy that combines multiple models through bagging, boosting, or stacking to achieve superior predictive performance and robustness.
Unsupervised learning and dimensionality reduction
- Implement clustering algorithms including k-means, hierarchical clustering, DBSCAN, and Gaussian mixture models with appropriate distance metrics and cluster validation.
- Apply dimensionality reduction techniques including PCA, t-SNE, UMAP, and autoencoders to visualize high-dimensional data and reduce feature space for downstream modeling.
- Analyze unsupervised learning results by evaluating silhouette scores, explained variance ratios, and domain-specific coherence measures to validate discovered patterns.
Model selection and optimization
- Implement hyperparameter optimization using grid search, random search, Bayesian optimization, and early stopping strategies to maximize model performance efficiently.
- Apply cross-validation strategies including k-fold, stratified k-fold, time series split, and nested cross-validation to obtain robust model performance estimates.
- Analyze model complexity trade-offs between underfitting and overfitting by evaluating learning curves, regularization strength, and generalization performance on holdout data.
- Design a model selection strategy that balances predictive accuracy, interpretability, inference speed, and maintenance complexity for production deployment decisions.
4
Domain 4: Deep Learning and Advanced Analytics
3 topics
Deep learning fundamentals
- Implement neural network architectures including feedforward networks, convolutional neural networks, and recurrent neural networks using deep learning frameworks.
- Apply training optimization techniques including learning rate scheduling, batch normalization, dropout regularization, and early stopping to prevent overfitting.
- Analyze deep learning model behavior using interpretability techniques including attention visualization, SHAP values, LIME explanations, and gradient-based saliency maps.
Natural language processing and specialized analytics
- Implement NLP pipelines including text preprocessing, tokenization, embedding representations, sentiment analysis, and named entity recognition for text analytics.
- Apply recommendation system techniques including collaborative filtering, content-based filtering, and hybrid approaches for personalized content delivery.
- Design a strategy for selecting between traditional ML, deep learning, and large language model approaches based on data availability, interpretability requirements, and computational constraints.
Graph analytics and optimization
- Apply graph analytics techniques including centrality measures, community detection, and link prediction for network analysis in social, biological, and financial domains.
- Implement optimization techniques including linear programming, constraint satisfaction, and genetic algorithms for resource allocation and scheduling problems.
- Analyze the appropriateness of different analytics techniques by evaluating data characteristics, problem structure, computational requirements, and solution quality trade-offs.
5
Domain 5: Model Deployment and MLOps
2 topics
Model deployment and serving
- Implement model deployment pipelines including model serialization, containerization, API serving, batch inference, and edge deployment strategies.
- Apply MLOps practices including model versioning, experiment tracking, automated retraining, A/B testing, and canary deployments for production ML systems.
- Analyze model performance degradation in production by monitoring data drift, concept drift, prediction quality, and latency to trigger retraining workflows.
- Design an end-to-end MLOps strategy that establishes CI/CD pipelines for ML, model governance workflows, and automated monitoring for production machine learning systems.
Feature stores and data platforms for ML
- Implement feature store architectures including online and offline stores, feature computation pipelines, feature versioning, and feature sharing across data science teams.
- Apply data platform engineering for ML including training data management, experiment tracking databases, model artifact storage, and serving infrastructure configuration.
- Design a data science platform strategy that standardizes tooling, enables reproducibility, supports collaboration, and scales compute resources for enterprise ML workloads.
6
Domain 6: Ethics, Governance, and Communication
3 topics
Responsible AI and data science ethics
- Implement bias detection and mitigation techniques including disparate impact analysis, fairness metrics, pre-processing debiasing, and post-processing calibration for ML models.
- Apply model documentation practices including model cards, datasheets for datasets, and impact assessments to ensure transparency and accountability in AI deployments.
- Analyze the ethical implications of data science solutions by evaluating fairness outcomes, privacy risks, potential for harm, and alignment with organizational values.
- Design a responsible AI governance framework that establishes ethical review processes, monitoring controls, and accountability structures for data science teams.
Data science communication and business impact
- Apply data visualization and storytelling techniques to communicate analytical findings, model results, and recommendations to non-technical business stakeholders.
- Analyze the business value of data science initiatives by quantifying revenue impact, cost savings, operational improvements, and risk reduction achieved through analytical solutions.
- Design a data science team operating model that defines roles, collaboration workflows, project intake processes, and success measurement aligned with organizational objectives.
AI governance and regulatory compliance
- Implement AI model risk management including model inventory, risk tiering, validation testing, and ongoing monitoring aligned with regulatory expectations for AI systems.
- Analyze regulatory compliance requirements for AI systems including the EU AI Act risk categories, algorithmic accountability mandates, and industry-specific AI regulations.
- Design an AI governance operating model that establishes model review boards, deployment approval workflows, and continuous monitoring for production AI systems.
Scope
Included Topics
- Data science methodology including CRISP-DM, the data science lifecycle, problem formulation, hypothesis development, experimental design, and result communication as tested on the ICCP Certified Data Scientist exam.
- Statistical foundations including descriptive statistics, probability distributions, hypothesis testing, confidence intervals, regression analysis, Bayesian inference, and experimental design.
- Machine learning techniques including supervised learning (classification, regression), unsupervised learning (clustering, dimensionality reduction), ensemble methods, model evaluation, and hyperparameter tuning.
- Advanced analytics including natural language processing, deep learning fundamentals, recommendation systems, time series analysis, and optimization techniques.
- Data engineering for data science including feature engineering, data preparation pipelines, model deployment, MLOps practices, and data science platform management.
- Ethical AI and responsible data science including bias detection, fairness metrics, model explainability, privacy-preserving techniques, and regulatory compliance for AI systems.
Not Covered
- Foundation-level data management fundamentals that are prerequisites for this certification and covered by the CDP Foundation exam.
- Big data infrastructure engineering including cluster management, distributed storage administration, and network optimization covered by the CBDP certification.
- Pure software engineering practices beyond what is necessary for data science model development and deployment.
- Academic-level mathematical proofs and theoretical computer science beyond the applied statistics and machine learning needed for the certification.
Official Exam Page
Learn more at ICCP
CDS is coming soon
Adaptive learning that maps your knowledge and closes your gaps.
Create Free Account to Be Notified