Machine Learning For Biomedical Applications
Module code: 6CCYB064
Module credits: 15
Module convenors: Dr Maria Deprez, Dr Emma Robinson
The module will provide the students with a fundamental grounding in the theoretical and computational skills required to apply machine learning tools to real-world problems. It will provide an understanding of the application of these skills to explore complex high-dimensional data sets; providing an overview of active research areas in machine learning, with biomedical applications.
On completion of the course the students should be able to:
- Have acquired a basic understanding of the most fundamental concepts related to machine learning
- Understand and be able to apply a range of commonly used machine learning techniques
- Have acquired practical computational skills that are needed to manipulate complex data sets
- Know how to practically design machine learning experiments
- Apply techniques learnt to domains related to biomedical applications
There will be 40 contact hours. This module emphasises hands-on tutorial sessions which will be a mix of lecture-style delivery followed immediately by problem solving activities.
- Introduction to Machine Learning for Biomedical data analysis.
- Introduction to programming in Python and Python packages: NumPy, Matplotlib, Pandas, scikit-learn.
- Fundamentals of Machine Learning: regression, classification and clustering; overfitting and cross-validation; performance measures.
- Linear, penalised, logistic and kernel regression, Gaussian process.
- Support vector machines, the kernel trick.
- Dimensionality reduction: Principal Component Analysis, Independent Component Analysis, Manifold learning
- Ensemble methods: Bagging, Boosting, Random Forest
- Feature selection
- Unsupervised methods: Gaussian Mixture Model, Markov Random Fields, Graph-cut algorithm
- Introduction to Deep Learning with TensorFlow: Multi-layer perceptron, Deep Convolutional Neural Networks.
Details of the module's summative assessment/s
Written exam (2 hours) (January)
2 coursework submissions
Formative assessment is done informally during lab sessions.
e-Learning: 6CCYB064 on KEATS
Pattern Recognition and Machine Learning, Christopher M. Bishop
Hands-On Machine Learning with Scikit-Learn and TensorFlow, Aurelien Geron