Lectures

  1. Lecture September 25, 2017 -  Introduction to the course. Exams and teaching material. Generalities on Machine Learning. (Ref. slide 1st-lecture)
  2. Lecture September 28, 2017 - The learning process. Data, learning machine. Minimization of the expected risk. Over and under fitting. Hint on Vapnik Chervonenkis theory and Structural risk minimization principle.  (Ref. slide 2nd lecture, Bishop – Pattern Recognition and Machine Learning, Springer, 2006 (Chap 1), V. Cherlassky, F. Mulier - Learning from Data, John Wiley and Sons, 2007 (chap.1 and 2) , Teaching Notes chapt. 1)
  3. Lecture October 2, 2017 -  Vc dimensione of oriented hyperplanes (theorem and proof) (slide). - The perceptron algorithm (Ref. Slide 3rd lecture, Teching notes chapt.2, P. Domingos - A few useful things to know about Machine Learning).
  4. Lecture October 5, 2017 - Python + libraries (Numpy, SkLearn, pandas, scipy.optimize) by Tommaso Colombo
  5. Lecture October 9, 2017 - The perceptron algorithm (proof). Brief review on optimization problems and optimization algorithm: basic defintions, convex problems, optimality conditions, unconstrained algorithms.
  6. Lecture October 12, 2017 - Beyond perceptron. Feedforward Neural Network: Multiplayer perceptron (Ref. Slide 6th Lecture, teaching notes chapt. 3)
  7. Lecture October 16, 2017 - The weights optimization problem. The full gradient method (ref. Slide of the 7th lecture)
  8. Lecture October 19, 2017 - Backpropagation rule - Convergence of the batch BP (learning rate) - Momentum term (Ref. Slide of 8th Lecture, teaching notes chapt. 3)
  9. Lecture October 23, 2017 -  Online methods for MLP: incremental, stochastic gradient, batch method (Ref. Slide of the 9th lecture; in depth reading:  Optimization methods for large-scale machine learning by L. Bottou, FE Curtis, J Nocedal)
  10. Lecture October 30, 2017 - Decomposition methods for MLP: Extreme learning and beyond (Ref. Slide of the 10-th lecture)
  11. Lecture November 2, 2017 - Regularized RBF; generalized RBF network; learning paradigm for RBF network: unsupervised versus supervised selection of the centers (Ref. Slide of the 11th lecture; Teaching notes chapt. 4)
  12. Lecture November 6, 2017 - Supervised selection of the centers Full optimization. Two block decompostion methods, exact and inexact solution of subproblems, convergence properties. Decomposition methods: block learning of centers (Ref. Slide of the 12th lecture; Teaching notes chapt. 4)
  13. Lecture November 9, 2017 - Introduction to TensorFlow by Tommaso Colombo
  14. Lecture November 13, 2017 - Hard SVM: generalities, defintion of margin, defintion of the max margin problem (Ref. Slide 14th lecture)
  15. MIDTERM November 16, 2017
  16. November 20, 2017 - -Hard SVM: generalities, defintion of margin, defintion of the max margin problem (Ref. Slide 11th lecture)
  17. November 23, 2017 - -Duality in convex Quadratic Programming
  18. November 27, 2017 - The dual problem of hard-SVM and C-SVM.
  19. November 30, 2017 - Tommaso Colombo
  20. December 4, 2017 - Soft non linear SVM: kernel defintion and examples
  21. December 11, 2017 SVM - Dual problem optimality conditions (KKT and feasible&descent directions)