Objectifs :
The aim of this lecture is to provide the essential knowledge to perform statistical data analysis and an introduciton to modern machine learning methods. These techniques are widely used in many scientific areas and their importance is continuously growing. Hands-on sessions will showcase applications implemented in Python and come as a support of the lectures to allow for the techniques to be easily and efficiently used afterwards. This course is addressed to all PhD students, with no specific knowledge of statistical analysis, machine learning or Python required.

Programme :
Part I – Statistical data analysis
• Fundamental concepts: probabilities, statistical model, likelihood function
• Statistical inference: parameter estimation, hypothesis tests, frequentist and bayesian approaches
• Hands-on session : data analysis, fitting a statistical model

Part II – Introduction to machine learning and neural networks
• Basic concepts: linear regression and classification, supervised and unsupervised learning
• Introduction to neural networks: constructing and training a network, review of popular neural networks used today
• Hands-on session: setting up a neural network using basic tools (python, numpy) and advanced machine learning libraries (pytorch)

Pré-requis :
This course is addressed to all PhD students of the EDSF, no specific knowledge of statistical analysis, machine learning or Python language is required.

Equipe pédagogique :
PASCUAL Bruna