This module will introduce students to the basic structure and operation of a typical data processing pipeline with a view towards enabling decision making. Students will be given the opportunity to start with a dataset and take it through all the stages needed to arrive at a decision using a data driven pipeline: data exploration, preprocessing, feature representation and engineering, finishing with decision making approaches. The module will emphasise the importance of selecting the correct approach to make appropriate decisions, introducing both supervised and unsupervised techniques, and the distinction between tasks such as classification, regression and clustering.
The module will avoid requiring the students to write code, instead using an existing data analytics framework. This is to introduce students to the concepts while their programming skills are still developing.
- Fundamentals
- The data analytics lifecycle
- Data analysis pipelines
- Data exploration and pre-processing
- Exploratory data analysis
- Data cleaning, pre-processing and profiling
- Feature selection and feature engineering
- Data transformations
- Decision-making techniques
- Regression vs. classification
- Supervised learning (classification, regression)
- Unsupervised learning (clustering, dimensionality reduction)
- Semi-supervised and reinforcement learning
- Model evaluation and validation
- Ensemble methods concepts
- Time Series Analysis
This is a one semester module, which will introduce students to a range of concepts that they will need to understand to pursue a career in data science and/or artificial intelligence. For this reason, the module delivery should emphasise an intuitive understanding over simple rote citation of definitions. This will be done using data analytics frameworks such as Orange Data Mining, which provide graphical or web-based interfaces that allow for visualisation and experimentation.
The module delivery will take the form of lectures and practical lab work. The labs will give students a practical opportunity to try the concepts discussed in the lectures. Students will submit lab assessments to encourage attendance and engagement. Students will also complete an individual assignment on a real-world dataset. This will involve picking a dataset, and pushing it through a data analytics pipeline to extract useful information.
| Module Content & Assessment | |
|---|---|
| Assessment Breakdown | % |
| Formal Examination | 50 |
| Other Assessment(s) | 50 |