MODULE 1: INTRODUCTION TO COMPUTATIONAL MODELLING FOR BI PROFESSIONALS
Topics:
- Introduction to advanced computational methods
- Discussion of analytic use cases
- Characteristics of statistical algorithm and data model requirements
- Assumptions and limitations of computational methods
Mapped to:
- K1 Types of algorithms and advanced computational methods
- K2 Range and application of various statistical algorithms
- K3 Range and application of various types of data models
- A1 Identify appropriate statistical algorithms and data models to test hypotheses or theories
- A3 Utilise a range of statistical methods and analytics approaches to data
Rationale for Sequencing of the Units
- The first learning unit introduces learners to advanced computational methods. Learners are taught the various types of statistical algorithms, the characteristics and assumptions and limitations of computational methods. There is a discussion of analytic use cases, e.g. diagnostic analytics, clustering, prediction, network analysis, text analytics, image analytics, video analytics, modelling and simulation.
MODULE 2: TECHNOLOGY STRATEGY
Topics:
- Trends in analytic development and production platform
- Technology strategy for development and operation of analytics
Mapped to:
- K4 Usage of analytics platforms and tools
- A2 Use appropriate analytics platforms and analytical tools given specific analytics and reporting requirements
Rationale for Sequencing of the Units
- Subsequently, the next learning unit teaches learners technology strategy by first going through trends in analytic development and platform strategy.
- The analytics platforms and analytical tools learners introduced to learners include R, SAS, Gretl, Orange.
MODULE 3: ADVANCED COMPUTATIONAL MODELLING DEVELOPMENT
Topics:
- Introduction to Statistical Modelling
- Introduction to Machine Learning Modelling
- Introduction to Time Series Forecasting Modelling
Mapped to:
- K5 Statistical modelling techniques
- A4 Conduct statistical modelling of data to derive patterns and/or solutions
- A6 Conduct tests on the actions taken and outcomes to assess effectiveness of the model
- A10 Draw relevant trends and insights from data analysis to support decisions
Rationale for Sequencing of the Units
- In the next learning unit, the learners move to the next step in the sequence where they are brought through statistical modelling techniques for machine learning, e.g.
- Supervised machine learning –decision tree
- Unsupervised machine learning –clustering
- Learners work on exercises to practice Decision Tre, K-mean clustering, hierarchical clustering.
Range of Application:
MODULE 4: ADVANCED COMPUTATIONAL MODELLING CODING
Topics:
- Data processing
- Data training and testing split
- Data sample balancing (optional)
- Model setting
- Model result report
Mapped to:
- K6 Coding languages for programming of algorithms and signals
- A5 Perform coding and configuration of software agents or programs based on a selected model or algorithm
- A8 Propose changes or updates to the model or algorithms applied
- A9 Implement changes to the coding and configuration of software agents or programs
Rationale for Sequencing of the Units
- In the next learning unit, the learners are taught data processing, data training and testing split, data sample balancing (optional), model setting and model result report.
- The software agents or programs learners use include Python, R, Gretl, JMP, SAS, Orange.
MODULE 5: POTENTIAL RISKS OF ANALYTICS MODEL
Topics:
- Unintended outcomes produced by analytical models and their potential reasons
Mapped to:
- K7 Potential reasons for unintended outcomes
- A7 Diagnose unintended outcomes produced by analytical models
Rationale for Sequencing of the Units
- The last learning unit teaches learners on potential risks of analytics models by looking at unintended outcomes, the potential reasons for unintended outcomes and how to diagnose various unintended outcomes.
