Back to ListMachine Learning
Lecturer
Ioannis Patras, i.patras@qmul.ac.uk
Content and organization
This module covers the following key concepts and themes:
The fundamentals:
- Introduction to Machine Learning
- Probability and Random Variables
Supervised Machine Learning Methods:
- Regression (Linear, no Linear, Multivariate)
- Classification I(Linear, no Linear, regulariston)
- Classification I (Decision Trees, Naïve bayes, metrics)
- Neural Networks
Unsupervised Machine Learning Methods:
- Clustering (k-means, hierarchical)
- Density Estimation (parametric distributions)
- Dimensionality reduction
Advanced Topics:
- Deep Learning, convolutional NN
- Ensembles
It covers methods for machine learning from signals and data, including statistical pattern recognition methods, neural networks, and clustering. During the module you will learn how to:
-
build predictive models from previous examples
-
organise your data in the case that no annotation of the data is available
-
evaluate machine learning methods
-
decide which learning model is most appropriate for the problem in hand
-
deal with situations when few learning examples are available.
You will also learn about fundamental methodologies that drove the Machine Learning field in the last decades, such as:
-
linear and nonlinear regression
-
classification with logistic regression and Bayesian Classifiers
-
clustering with k-means algorithm
-
density estimation and dimension reduction
In addition, you will learn about state of the art methodologies, such as Deep Neural Networks and ensemble methods that are behind the recent renaissance of AI and have found applications in face detection and recognition, speech recognition and medical diagnosis, among others.
Course Type
AI PhD Curriculum
