Fall Semester


Matrix Computation


Former lecturer of  this course was Stéphane Canu.

Overview

  • Introduction to matrix calculus
  • Notion of numerical complexity
  • Matrix Factorizations
    • LU
    • LDL and Cholesky factorization
    • QR factorization
  • Solving  linear system
    • Review of Least squares problem
    • Application of matrix factorizations
    • Iterative methods (Jacobi, Gauss-Seidel, relaxation)
  • Algorithms for Eigenvalues and Singular values computation
Complete materials of this course are available here.

Spring Semester


Statistical Learning

Overview

This course aims at introducing to machine learning wih a main focus on linear decision function learning.

  • Introduction to data-mining and statistical learning 
  • Principal Component Analysis 
  • Clustering methods
  • Theory of Bayes Decision
  • Linear regression for classification
  • Logistic Regression
  • Support Vector Machine
More materials about the course are available here

Dynamical systems and Linear Control

Overview

The course aims at introducing the basic notions of analyzing linear dynamical systems in time and frequency domain. A step is taken further to address the notion of feedback, analysis of system stability in closed-loop and design of linear controllers (PID). Another part of the course introduces state-space modelling approach, its properties and design of state-space feedback and observers.

  • Dynamical Linear Invariant Time system, transfer function, block diagram algebra
  • Time domain and Frequency domain analysis
  • Notion of feedback and the need of closed-loop
  • Stability analysis, stability robustness
  • Transient performances, steady state error
  • Design of linear controllers (PID)
  • Stae-space modelling
  • Controllability, observability
  • Design of state-space feedback and observers
Complete materials of the course are available here

Projects


Intelligent Housing


We are developing a platform of Intelligent housing with the objective to adapt automatically to the changing needs of residents. The platform incorporates sensors, remote control lights and plugs, usual home electrical appliances. The pursued goal is collecting data from the platform and inferring the residents' activities and habits using statistical tools.  Feedbacks to the users and interactions with the systems are intended via Multi-Agent Systems.


Humanitarian Project M'Bolo

The challenge of the project is to popularize the use of Computers and handy softwares in African countries. For this sake, students of my department collect and haul first or second hand materials (computers, printers, overhead projectors, books, free softwares, ...) to high schools in these countries. They also organized two weeks Spring School for active and intensive training in the high schools.