Advanced Course in Machine Learning : Lectures

The course consists of 13 lectures. The upcoming lecture topics are somewhat tentative and subject to minor changes. The lecture slides will be added here after the lectures.

Part I: Preliminaries and general conecpts

  1. Tuesday 15.3.: Introduction, practicalities and prerequisities. Includes quick summary of probabilites and linear algebra.
    1. [SLIDES]
    2. Alternative reading:  The UML lecture notes section 2
  2. Thursday 17.3.: Machine learning basics: Loss functions, overfitting, regularization, and all that. The other half covers convex optimization.
    1. [SLIDES]
    2. Alternative reading: The UML lecture notes section 3
  3. Tuesday 22.3.: Learning tasks and probabilistic models.
    1. [SLIDES]
    2. Alternative reading: David Barber's book section 8.7, 8.8, parts of Sections 4, 9 and 13

Part II: Unsupervised learning

  1. Thursday 31.3.: Clustering: spectral clustering, mixture models and the EM algorithm
    1. [SLIDES]
    2. Alternative reading:  The UML lecture notes section 12
  2. Tuesday 5.4.: Linear dimensionality reduction: PCA, factor analysis, ICA
    1. [SLIDES]
    2. Alternative reading: The UML lecture notes sections 4-8
  3. Thursday 7.4.: Matrix factorization and non-linear dimensionality reduction
    1. [SLIDES]

Part III: Supervised learning

  1. Tuesday 12.4.: Linear supervised models for regression and classification; regression and sparse linear models; generative vs discriminative classifiers
    1. [SLIDES]
  2. Thursday 14.4.: Kernel methods and support vector machines
    1. [SLIDES]
  3. Tuesday 19.4.: Decision trees, boosting and ensembles
    1. [SLIDES]

Part IV: Neural networks and deep learning

  1. Thursday 21.4: Neural network concepts, multilayer perceptron, backpropagation
    1. [SLIDES]
    2. Alternative reading: Selected parts of the Deep learning book by Goodfellow, Bengio and Courville
  2. Tuesday 26.4.: Deep learning: convolutional neural networks and supervised deep learning
    1. [SLIDES]
  3. Thursday 28.4.: Unsupervised deep learning: autoencoders and Boltzmann machines
    1. [SLIDES]

Part V: Re-cap and exam preparation

  1. [SLIDES]
Tulostettava sivuTulostettava sivu