Artificial Intelligence and Deep Learning in Healthcare

Course ID   MSC05

Semester   B

Type   Compulsory

The course of Artificial Intelligence and Deep Learning in Healthcare aims at teaching fundamental concepts of Artificial Intelligence and Machine Learning, and their applications in healthcare data. It focuses on tabular data, while it explores several Machine Learning models, including Deep Learning with Artificial Neural Networks. Upon successfully completing this course, students will have basic theoretical knowledge around Machine Learning and its healthcare applications, as well as hands-on experience with modern software frameworks including scikit-learn and Tensorflow/Keras.


Introduction to Artificial Intelligence and example applications in Healthcare.

Basic principles and concepts of Machine Learning - Types of problems, experience, tasks, performance measures

Simple and multiple linear regression and normal equations 

Introduction to scikit-learn

Linear regression and maximum likelihood estimation

Data preparation: Normalization, standardization, encoding of categorical variables, encoding of circular variables

Introduction to Multilayer Artificial Neural Networks (Multi-Layer Perceptrons)

Hidden layers and activation functions

Parameter initialization and loss functions

Overview of ANN training and the backproparagion algorithm

Introduction to Tensorflow/Keras

Capacity and generalization in machine learning problems

Overview of regularization and capacity control

Bayesian methods for classification 

Probability density estimation algorithms 

k-NN classification

Project including application to healthcare datasets