Skip to main content

Applied Machine Learning

Module information>

Academic Direction
Birkbeck, University of London
Modes of Study

Machine learning is an important topic in both academia and industry these days. There has been growing interest in the practical side of machine learning.

This module focuses more on the practical techniques and methods with Python and Scikit-Learn than on the theories or statistics behind these methods.

You will gain hands-on and practical skills for machine learning based analytics tasks, use appropriate Python libraries and tools to analyse data, and develop the design and programming skills that will help build intelligent artefacts.

The module helps you assess the performance of machine learning models and develop a deeper understanding of several real-life topics in applied machine learning, in order to develop the practical skills necessary to pursue research in applied machine learning.

Topics covered

Main topics of the module include:

  • Introduction to Python for machine learning
  • Preparing data
  • Feature selection for machine learning
  • Resampling
  • Feature evaluation
  • Rule-based algorithms: decision tree and random forest
  • Regression-based algorithms: logistic regression and neural networks
  • Large-scale machine learning using TensorFlow
  • Real-life case studies: financial forecasting
  • Real-life case studies: computer vision.

Learning outcomes

Upon successful completion of this module, you will be able to:

  • Apply machine learning tools to solve practical problems in real-life scenarios.
  • Evaluate and identify appropriate machine learning methods and techniques to analyse data.
  • Critically analyse and interpret machine learning results.
  • Demonstrate deep understanding of a range of complex real-life topics in applied machine learning.
  • Evidence an awareness of the steps involved in constructing machine learning architectures.
  • Demonstrate an awareness of approaches for avoiding poor model performance, including under and over fitting of the model to the training data.


Two assessment components:

  • Online auto-graded test (25%)
  • End of term coursework (75%)