CPDS Certification Specification | Data Science Institute

Certification specification

Certified Professional Data Scientist

The Certified Professional Data Scientist (CPDS®) is the Data Science Institute's advanced professional credential in applied data science.

CPDS certification page

CPDS specification at a glance
Level	Professional
Structure	4 Parts, 12 Units
Assessment	Exam A, then Exam B
Pass mark	50% in each exam
Exam delivery	Online, remotely proctored
Builds on	CPDA®

Section 1Certification overview

The Certified Professional Data Scientist (CPDS®) is the Data Science Institute's advanced professional credential in applied data science. It validates competence across the data science lifecycle, including data preparation, advanced analytical methods, machine learning, neural networks and deep learning foundations.

CPDS® develops applied capability using Python-based workflows, realistic datasets, worked examples and practice activities that reflect professional data science tasks. The certification places strong emphasis on method selection, model interpretation, evaluation, communication and responsible applied practice.

CPDS® is designed for learners with prior analytics, statistics or programming experience. Its audiences include professional data analysts progressing beyond CPDA®, technical professionals moving into data science roles, graduates with quantitative backgrounds, and practitioners who need formal validation of applied machine learning and advanced data science capability.

CPDS® sits above CPDA® on the DSI pathway and is intentionally more advanced in analytical and technical scope. It assumes that learners can prepare data, interpret statistical outputs and work in Python before moving into advanced data science methods, machine learning and deep learning concepts.

The certification is delivered online through the Certifications Platform, with structured learning activities and two summative certification examinations delivered online with remote proctoring.

Section 2Programme structure

2.1 Structural model

The programme is organised as Parts → Units → Lessons, with unit scope defined in the unit-level syllabus and lesson-level descriptors defined in the lesson-by-lesson breakdown.

2.2 Parts and weightings

Parts and weightings for CPDS
Part	Share of the certification	Weighting
Data Scientist FoundationPart 1		10%
Advanced Data Science MethodsPart 2		40%
Machine LearningPart 3		40%
Neural Networks and Deep LearningPart 4		10%

2.3 Parts and units

Part 1 · 10%

Data Scientist Foundation

Unit 1 Data Management in Python
Unit 2 Data Visualisation in Python
Unit 3 Inferential Parametric Tests
Unit 4 Predictive Modelling and Data Reduction Foundations

Part 2 · 40%

Advanced Data Science Methods

Unit 1 Time Series Analysis
Unit 2 Advanced High-Dimensional Data Analysis
Unit 3 Natural Language Processing

Part 3 · 40%

Machine Learning

Unit 1 Foundations of Supervised Learning
Unit 2 Decision Trees and Ensemble Learning
Unit 3 Advanced Models and Specialised Techniques

Part 4 · 10%

Neural Networks and Deep Learning

Unit 1 Neural Networks
Unit 2 Deep Learning

2.4 Programme scale

This consolidated specification includes 4 Parts and 12 Units. The certification is supported by lesson learning materials, knowledge checks, unit practices, and part-level case studies delivered through the Certifications Platform.

Section 3Learning and assessment requirements

3.1 Standard learning pathway

The programme combines instruction, applied practice, and objective checks at lesson level; consolidation at unit level; part-level applied case work where present; and certification readiness activities supported by two summative certification examinations at certification level.

3.2 Progression requirements

To be eligible for award:

All Unit Practices must be completed (ungraded but mandatory).
At least 90% of platform learning activities must be marked complete, including applied activities and case studies where present in the pathway.
Both CPDS® summative certification examinations must be passed.

3.3 Summative certification exams

Exam A — Core Knowledge Exam

Type: Knowledge
Format: Computer-based, objective examination
Duration: 180 minutes
Pass mark: 50%
Sequence: Must be passed before Exam B can be attempted

Exam B — Practical Application Exam

Type: Practical application
Format: Applied practical examination
Duration: 180 minutes
Pass mark: 50%
Sequence: Taken after Exam A is passed

Assessment delivery environment is online with remote proctoring.

CPDS® certification exam structure
Attribute	Exam A — Core Knowledge Exam	Exam B — Practical Application Exam
Type	Knowledge	Practical application
Format	Computer-based, objective examination	Applied practical examination
Duration	180 minutes	180 minutes
Pass mark	50%	50%
Sequence	Must be passed before Exam B can be attempted	Taken after Exam A is passed

3.4 Use of the Certifications Platform

CPDS® is delivered through the Institute's Certifications Platform, which structures each certification consistently. Each Part contains units, units contain lessons, and supporting practice and case-study material is provided at the appropriate level of the structure.

At Part level — a Case Study integrating the units within the Part, where appropriate to the structure and learning outcomes.
At Unit level — a Unit Practice that demonstrates capability against the unit's ILOs.
At Lesson level — Learn content, Lesson Practice, and a Knowledge Check aligned to the lesson's ILOs.
Across the programme — curated datasets, worked examples, Python workflows, modelling activities and quizzes that support lesson-level practice, applied consolidation and revision for the certification examinations.

Section 4Part descriptors

Each part below sets out its scope, its intended learning outcomes, and the units it contains. Unit entries give the unit purpose, the applied outputs expected, the tools and environment used, the scope of the unit, and the unit intended learning outcomes.

4.1 Part 1 · 10% of the certificationData Scientist Foundation

Part 1 establishes foundational competence for CPDS®: practical Python data management, exploratory visual analysis, introductory inferential testing, baseline predictive modelling, and essential data reduction concepts required for advanced data science methods. It provides a common foundation for learners entering CPDS® from different academic or professional backgrounds, including learners who have not completed CPDA®.

Part intended learning outcomes

Prepare and manage datasets in Python for analysis and modelling using reproducible workflows.
Create and interpret exploratory visualisations to understand patterns and relationships in data.
Apply introductory parametric hypothesis tests and interpret outputs and assumptions at a foundational level.
Fit and interpret baseline regression and logistic regression models for simple predictive tasks.
Explain foundational dimensionality reduction and clustering concepts required for progression into advanced high-dimensional data analysis.

Unit 1

Data Management in Python

Unit purpose: Develop practical competence in preparing and managing datasets for analysis and modelling, including import/export, inspection, transformation, aggregation, missing data handling and ETL awareness.
Applied outputs: Cleaned, analysis-ready datasets with reproducible preparation steps and an ability to describe the role of ETL in analytics pipelines.
Tools and environment: Python with pandas and associated core libraries introduced in a notebook environment.
Unit scope: Import/export and data inspection; feature modification; subsets and sorting; combining datasets and aggregation; strategies for missing values; and ETL awareness and pipeline framing.

Unit intended learning outcomes

Import and export data from common formats and sources using reproducible Python workflows.
Inspect datasets to identify data-quality issues and apply appropriate cleaning and modification techniques.
Create subsets, filter records and sort data to prepare analysis-ready views.
Combine datasets and compute grouped summaries using merging, appending and aggregation operations.
Detect and handle missing values using methods introduced, documenting assumptions and impacts.
Explain the purpose of ETL and describe how ETL steps relate to analytics project workflows.

Unit 2

Data Visualisation in Python

Unit purpose: Build practical capability to create and interpret common analytical charts in Python for exploratory analysis and communication.
Applied outputs: Reproducible visual outputs demonstrating appropriate chart selection and clear interpretation of patterns, comparisons and relationships.
Tools and environment: Python visualisation environment using common charting libraries as introduced in the lessons.
Unit scope: Core chart types for comparison and distribution; formatting for readability; and relationship visualisation for exploratory insight generation.

Unit intended learning outcomes

Create and format common chart types in Python for categorical comparison and distribution exploration using the methods introduced.
Select appropriate chart types for a given analytical question and justify the choice at a foundational level.
Visualise and interpret relationships between variables using relationship plots introduced.
Apply presentation best practices, including labels, titles, scales and layout, to improve clarity and interpretability.

Unit 3

Inferential Parametric Tests

Unit purpose: Introduce core parametric hypothesis tests in Python and develop competence in interpreting test outputs and assumptions at a foundational level.
Applied outputs: Executed parametric tests with clear interpretation of results and assumptions, including correct reporting of hypotheses and statistical conclusions.
Tools and environment: Python statistical testing tooling and workflows introduced in the lessons.
Unit scope: Introductory parametric testing workflows in Python, including selecting an appropriate test for common scenarios and interpreting outputs.

Unit intended learning outcomes

State null and alternative hypotheses for common parametric test scenarios and interpret p-values at a foundational level.
Select and run parametric hypothesis tests introduced in Python, checking assumptions where required.
Interpret test outputs and draw appropriate conclusions in context, including limitations and assumptions.

Unit 4

Predictive Modelling and Data Reduction Foundations

Unit purpose: Develop introductory predictive modelling capability through regression and binary logistic regression concepts and workflows, while also introducing the foundational data reduction and clustering concepts required for progression into advanced high-dimensional data analysis.
Applied outputs: Baseline regression and logistic regression models with interpreted coefficients and introductory performance interpretation; foundational PCA and K-Means interpretation outputs used as preparation for advanced high-dimensional workflows.
Tools and environment: Python modelling environment and libraries introduced in the lessons, including introductory preprocessing and dimensionality reduction workflows.
Unit scope: Introductory modelling workflow and interpretation for multiple linear regression and binary logistic regression; essential data reduction concepts including high-dimensional data, feature redundancy, multicollinearity, scaling, PCA foundations, K-Means foundations, and the distinction between supervised and unsupervised learning.

Unit intended learning outcomes

Fit and interpret baseline multiple linear regression models using methods introduced, including coefficient interpretation.
Fit and interpret baseline binary logistic regression models using methods introduced, including probability and odds interpretation at a foundational level.
Explain the difference between regression and classification modelling goals and relate model choice to outcome type.
Evaluate baseline model outputs using introductory metrics and interpretation approaches introduced.
Explain why dimensionality reduction is used in data science workflows.
Describe the practical role of PCA in reducing feature complexity and summarising correlated variables.
Interpret variance explained, component scores and component loadings at a foundational level.
Explain the purpose of clustering and distinguish it from supervised predictive modelling.
Describe the basic intuition of K-Means clustering and its use in segmentation.

4.2 Part 2 · 40% of the certificationAdvanced Data Science Methods

Part 2 develops advanced analytical methods used in professional practice: time series analysis and forecasting, advanced high-dimensional data analysis, and natural language processing. The part moves beyond foundational data reduction by introducing nonlinear dimensionality reduction, latent factor methods, density-based clustering, feature selection and regularised regression methods used in complex modelling contexts.

Part intended learning outcomes

Analyse time series data by identifying key components and selecting appropriate forecasting strategies.
Assess stationarity and apply differencing and diagnostic methods introduced to prepare time series for ARIMA-family modelling.
Build and interpret ARIMA and seasonal ARIMA models and use them for forecasting with appropriate diagnostics.
Evaluate high-dimensional datasets and select appropriate advanced reduction, representation, clustering or feature-selection methods.
Apply nonlinear dimensionality reduction methods introduced, including t-SNE and UMAP, and interpret reduced-dimensional outputs appropriately.
Apply factor analysis and interpret latent factor structure in applied data science contexts.
Apply density-based clustering methods introduced, including DBSCAN, and interpret cluster and noise assignments.
Apply feature-selection and regularised regression methods introduced to support high-dimensional predictive modelling.
Apply foundational NLP and text-mining steps introduced to convert raw text into analysis-ready representations and extract insights.

Unit 1

Time Series Analysis

Unit purpose: Develop applied capability in time series analysis and forecasting, including decomposition, stationarity concepts and ARIMA-family modelling.
Applied outputs: Time series decompositions and diagnostic plots; stationarity assessment outputs; fitted ARIMA/SARIMA models; and forecast outputs with interpreted diagnostics.
Tools and environment: Python time series analysis and modelling workflow using the libraries and procedures introduced in the lessons.
Unit scope: Time series concepts; decomposition; stationarity and transformations; ARIMA modelling; seasonal ARIMA extensions; forecasting; and interpretation of outputs and diagnostics.

Unit intended learning outcomes

Describe key characteristics of time series data and distinguish components such as trend, seasonality and residual variation.
Perform time series decomposition and interpret component outputs to inform modelling choices.
Assess stationarity and apply stationarity-preparation steps introduced, including differencing, where appropriate.
Build ARIMA models and interpret parameters and outputs at an applied level.
Build seasonal ARIMA models and use them to generate forecasts, interpreting results and diagnostics introduced.

Unit 2

Advanced High-Dimensional Data Analysis

Unit purpose: Develop capability to analyse complex, high-dimensional datasets using advanced representation, clustering, feature-selection and regularised modelling methods.
Applied outputs: t-SNE and UMAP embeddings; factor analysis outputs; DBSCAN cluster and noise assignments; feature-selection summaries; Lasso regression workflows; and interpreted high-dimensional analysis outputs linked to modelling and insight-generation decisions.
Tools and environment: Python-based high-dimensional data analysis workflow using pandas, scikit-learn and relevant dimensionality reduction, clustering and modelling libraries introduced in the lessons.
Unit scope: High-dimensional data challenges; nonlinear dimensionality reduction using t-SNE and UMAP; latent factor methods; density-based clustering with DBSCAN; feature selection in high-dimensional modelling; regularisation; and Lasso regression for shrinkage and feature selection.

Unit intended learning outcomes

Assess the challenges associated with high-dimensional datasets, including sparsity, redundancy, noise, multicollinearity and interpretability.
Apply t-SNE to create low-dimensional representations and interpret outputs with appropriate caution.
Apply UMAP to produce reduced-dimensional embeddings and explain its role in exploratory analysis and structure discovery.
Apply factor analysis to identify latent constructs and interpret factor loadings and factor structure in context.
Apply DBSCAN and interpret dense clusters, sparse regions and noise points.
Select and apply feature-selection methods for high-dimensional predictive modelling workflows.
Apply Lasso regression and explain how regularisation supports shrinkage, variable selection and model generalisation.
Compare advanced data reduction, clustering and feature-selection approaches and select suitable methods for applied data science problems.
Communicate the limitations and interpretation risks of high-dimensional analytical outputs clearly and responsibly.

Unit 3

Natural Language Processing

Unit purpose: Introduce text-mining and NLP foundations for converting unstructured text into analysis-ready representations and extracting insights from text data.
Applied outputs: Text preprocessing outputs; basic text representations introduced; and interpreted results from a simple text mining or NLP workflow.
Tools and environment: Python text-mining and NLP workflow using libraries and methods introduced in the lessons.
Unit scope: Text preprocessing and feature extraction; introductory NLP concepts and workflows; and interpretation of outputs in applied contexts.

Unit intended learning outcomes

Describe a basic text-mining workflow from raw text to analysis-ready features.
Apply preprocessing steps introduced and explain their purpose.
Create basic text representations introduced and use them to support simple analysis tasks.
Explain introductory NLP concepts and interpret outputs of an NLP workflow at a foundational level.

4.3 Part 3 · 40% of the certificationMachine Learning

Part 3 develops applied machine learning capability across supervised learning foundations, tree-based and ensemble methods, and specialised techniques used in professional contexts.

Part intended learning outcomes

Select appropriate supervised learning methods for classification problems and explain basic method assumptions and trade-offs at an applied level.
Build and evaluate baseline classification models using algorithms introduced and interpret model outputs.
Build and interpret decision tree models and relate tree structure to prediction rules.
Build and interpret random forest models and explain ensemble intuition at a foundational level.
Apply support vector machines using the workflow introduced and interpret results at a foundational level.
Apply market basket analysis concepts introduced and interpret frequent itemsets and rules for practical insight.
Apply WoE and information value concepts introduced for feature selection and interpret outputs in a modelling context.

Unit 1

Foundations of Supervised Learning

Unit purpose: Introduce supervised learning foundations through baseline classification algorithms and applied evaluation concepts.
Applied outputs: Fitted baseline classifiers and interpreted predictions; initial model performance summaries using evaluation concepts introduced.
Tools and environment: Python machine learning workflow using the libraries and procedures introduced in the lessons.
Unit scope: Classification framing; k-nearest neighbours workflow and interpretation; naive Bayes workflow, including probability framing and interpretation.

Unit intended learning outcomes

Explain the supervised classification task and distinguish features and target variables.
Fit and interpret a k-nearest neighbours classifier and explain how choice of k affects behaviour.
Fit and interpret naive Bayes classifiers, including basic probability interpretation as presented.
Evaluate baseline classification performance using measures and interpretation approaches introduced.

Unit 2

Decision Trees and Ensemble Learning

Unit purpose: Develop applied capability in tree-based modelling and ensemble learning using decision trees and random forests.
Applied outputs: Fitted decision tree and random forest models with interpreted outputs and performance summaries for practical classification tasks.
Tools and environment: Python machine learning workflow using the libraries and procedures introduced in the lessons.
Unit scope: Decision tree modelling; random forest modelling; ensemble concepts; applied workflow; and comparison of tree-based models using evaluation approaches introduced.

Unit intended learning outcomes

Fit and interpret decision tree models and explain how splits form decision rules.
Apply tree configuration concepts introduced and interpret impacts on performance.
Fit and interpret random forest models and explain ensemble intuition at a foundational level.
Compare tree-based models using evaluation approaches introduced and select a reasonable model for a practical task.

Unit 3

Advanced Models and Specialised Techniques

Unit purpose: Introduce specialised techniques used in professional data science contexts: market basket analysis, support vector machines, and feature selection using WoE and information value.
Applied outputs: Market basket analysis outputs; fitted SVM outputs; and WoE/IV feature-selection summaries with interpreted results.
Tools and environment: Python machine learning workflow using the libraries and procedures introduced in the lessons.
Unit scope: Association-rule and market basket analysis; SVM classification and interpretation; and WoE and information value for feature selection and model preparation.

Unit intended learning outcomes

Describe the purpose of market basket analysis and interpret itemsets and association rules introduced.
Fit and interpret a support vector machine classifier and describe decision boundaries at a foundational level.
Compute and interpret WoE and information value measures and use them to inform feature selection decisions.

4.4 Part 4 · 10% of the certificationNeural Networks and Deep Learning

Part 4 introduces neural networks and deep learning concepts and applied workflows. Learners develop understanding of neural network architecture and learning mechanisms and build baseline neural network models. The part also introduces deep learning architectures and application areas.

Part intended learning outcomes

Explain foundational neural network concepts and architecture, including neurons, layers and learning via backpropagation.
Build a baseline neural network model using the workflow introduced and interpret training outputs at a foundational level.
Describe deep learning motivation, common architectures and application areas introduced, such as computer vision and language applications.

Unit 1

Neural Networks

Unit purpose: Introduce the fundamentals of neural networks, including core concepts, architecture and learning mechanisms, and develop an applied model-building workflow.
Applied outputs: A baseline neural network model in Python, including data preparation or normalisation, model definition, training and evaluation outputs.
Tools and environment: Python neural network workflow using the libraries and approaches introduced in the lessons.
Unit scope: Neural network foundations; learning concepts introduced, including backpropagation; and a practical workflow for building a baseline neural network model.

Unit intended learning outcomes

Describe the key components of a neural network and explain how information flows through layers.
Explain learning mechanisms introduced, including backpropagation, at a foundational level.
Prepare data for a neural network model using normalisation or scaling steps introduced.
Build and train a baseline neural network model and interpret training outputs and basic evaluation results.

Unit 2

Deep Learning

Unit purpose: Introduce deep learning as an extension of neural networks and describe common deep learning architectures and application areas.
Applied outputs: Interpretation of deep learning architectures and applications, connecting architecture choices to task requirements at a foundational level.
Tools and environment: Conceptual and applied Python workflow elements as introduced in the lesson materials.
Unit scope: Deep learning motivation; deep architectures and representation learning; application domains such as computer vision and language processing; and high-level architectural concepts introduced.

Unit intended learning outcomes

Explain what distinguishes deep learning from shallow neural networks and describe the role of multiple layers for feature learning.
Describe deep learning architectures introduced and connect them to typical application areas, including vision and language.
Interpret a deep learning application scenario and explain why deep learning is appropriate for the task at a foundational level.

Section 5Appendices

Appendix A Position in the DSI Certification Pathway

CPDS® is the advanced professional data scientist credential in the DSI certification pathway. It builds on the applied analytical competence of CPDA® and moves learners into advanced methods, machine learning and deep learning foundations.

Learners who have not completed CPDA® should have equivalent preparation in Python data management, visual analysis, inferential statistics, predictive modelling foundations and introductory data reduction before attempting CPDS®. Part 1 provides a common foundation but does not replace the full breadth of CPDA® preparation.

CPDS certification page CPDA specification

Appendix B Platform Learning Terms

Lesson Practice refers to lesson-level applied activity that helps learners practise the concept or workflow immediately after learning it. Knowledge Checks are objective checks aligned to lesson learning outcomes. Unit Practices are mandatory consolidation activities linked to the unit's intended learning outcomes. Part-level Case Studies integrate learning across units and support applied professional judgement.