/AMAL

Advanced Machine Learning And Deep Learning

Primary LanguagePython

AMAL: Advanced MAchine Learning & Deep Learning

The aim of this course is to cover the following concepts:

Introduction to deep learning: Course and experimentations on state of the art architectures.

  • Deep neural Network architectures
  • Convolutional neural network, reccurents neural network.
  • Attention models.
  • Computational graph and autograd.
  • Framework for deeplearning: pytorch, tensor flow and CUDA programmation.

Deepening of the founding concepts of the machine learning

  • Statistical learning theory, generalizability, bias-variance dilemma, PAC, learning complexity, etc.
  • Supervised Learning: Classification, Neural Networks, Support Vector Machines, Kernel Methods, Gaussian Processes, etc...
  • Optimization
  • Unsupervised learning: Clustering, Matrix Factoring, Latent Variable Models (blends, etc.)

Other learning paradigms :

  • Low supervision learning, Semi-supervised and transductive learning, Active learning,
  • Transfer Learning
  • Learning and structured data: Sequences and trees, Graphs and interdependent data.