Section outline

◄Probabilistic (Generative) Learning (20h)
►Generative Deep Learning (8h)

  • The module presents the fundamental concepts, challenges, architectures and methodologies of deep learning. We introduce the learning of neural representations from vectorial, sequential and image data, covering both supervised and unsupervised learning, and hinting at various forms of weak supervision.  Models covered include: deep autoencoders, convolutional neural networks, long-short term memory, gated recurrent units, advanced recurrent architectures, sequence-to-sequence, neural attention, Transformers, neural Turing machines. Methodological lectures will be complemented by introductory seminars to Keras-TF and Pytorch.

    Date Topic   References  (OLD)
    		 References   (NEW) Additional Material 
    15
    		 27/03/2024
    (16-18)
    		 Convolutional Neural Networks I
    Introduction to the deep learning module; i    ntroduction to CNN; basic CNN elements
    		 [DL] Chapter 9
    		 [CHB] Chapter 10
    [SD] Chapter 10
    		Additional Readings
    [15-19] Original papers for LeNet, AlexNet, VGGNet, GoogLeNet and ResNet.
    16
    		 28/04/2024
    (14-16)    		  Convolutional Neural Networks II
    CNN architectures for image recognition; convolution visualization; advanced topics (deconvolution, dense nets); applications and code
    		  [DL] Chapter 9 [CHB] Chapter 10
    [SD] Chapter 10    		Additional Readings
    [20] Complete summary of convolution arithmetics
    [21] Seminal paper on batch normalization
    [22] CNN interpretation using deconvolutions
    [23] CNN interpretation with GradCAM


    		 EASTER BREAK


    17
    		 03/04/2024
    (16-18)
    		 Deep Autoencoders
    S    parse, denoising and contractive AE; deep RBM
    		 [DL] Chapter 14, Sect 20.3, 20.4.0 (from 20.4.1 onwards not needed)
    		 [CHB] Section 19.1

    [SD] Coverage of the Prince book on this lecture is inadequate.
    		Additional Readings
    [24] DBN: the paper that started deep learning
    [25] Deep Boltzmann machines paper
    [26] Review paper on deep generative models
    [27] Long review paper on autoencoders from the perspective of representation learning
    [28] Paper discussing regularized autoencoder as approximations of likelihood gradient
    18 04/04/2024
    (14-16)    		 Gated Recurrent Networks I
    Deep learning for sequence processing; gradient issues;
    		[DL] Sections 10.1-10.3, 10.5-10.7, 10.10, 10.11
    		 Coverage of the Bishop and Prince books on this lecture is inadequate (for reasons I do not understand). Please use the DL book or slides integrated by the Additional Readings.
    		Additional Readings
    [29] Paper describing gradient vanish/explosion
    [30] Original LSTM paper
    [31] An historical view on gated RNN
    19
    		 05/04/2024
    (16-18)
    ROOM E
    		 Gated Recurrent Networks II
    long-short term memory; gated recurrent units; generative use of RNN
    RECOVERY LECTURE
    		 [DL] Sections 10.12, 12.4.5
    		 Coverage of the Bishop and Prince books on this lecture is inadequate (for reasons I do not understand). Please use the DL book or slides integrated by the Additional Readings.
    		Additional Readings
    [32] Gated recurren units paper
    [33] Seminal paper on dropout regularization

     Software
    20
    		 09/04/2024
    (11-13)
    		 Coding practice I
    Pytorch and principles of autograd
    Guest lecture by Valerio De Caro


    21
    		 10/04/2024
    (16-18)
    		 Coding practice II
    Keras/TF and programming exercises
    Guest lecture by Valerio De Caro

     Github with the notebooks for the lecture: https://github.com/vdecaro/intro-tf-keras/


    22
    		 11/04/2024
    (14-16)
    		 Attention-based architectures
    sequence-to-sequence;  attention modules; transformers
    		  [DL] Sections 10.12, 12.4.5
    		 [CHB] Chapter 12
    [SD] Chapter 12
    		Additional Readings
    [34,35] Models of sequence-to-sequence and image-to-sequence transduction with attention
    [36] Seminal paper on Transformers
    [37] Transformers in vision
    23
    		 12/04/2024
    (16-18)
    AULA E
    RECOVERY LECTURE
    		 Memory-based models
    multiscale network; hierarchical models; memory networks; neural Turing machines

































◄Probabilistic (Generative) Learning (20h)
►Generative Deep Learning (8h)