Teaching plan for the course unit

 

Close imatge de maquetació

 

Print

 

General information

 

Course unit name: Deep Learning

Course unit code: 572669

Academic year: 2017-2018

Coordinator: Jordi Vitria Marca

Department: Department of Mathematics and Computer Science

Credits: 3

Single program: S

 

 

Popular Datasets

 

  •   ImageNet: Large-scale object dataset
  •   Microsoft Coco: Large-scale image recognition, segmentation, and captioning dataset
  •   Mnist: handwritten digits
  •   PASCAL VOC: Object recognition dataset
  •   KITTI: Autonomous driving dataset
  •   NYUv2: Indoor RGB-D dataset
  •   LSUN: Large-scale Scene Understanding challenge
  •   VQA: Visual question answering dataset
  •   Madlibs: Visual Madlibs (question answering)
  •   Flickr30K: Image captioning dataset
  •   Flickr30K Entities: Flick30K with phrase-to-region correspondences
  •   MovieDescription: a dataset for automatic description of movie clips
  •   Action datasets: a list of action recognition datasets
  •   MPI Sintel Dataset: optical flow dataset
  •   BookCorpus: a corpus of 11,000 books

 

 

Software

 

  •   Caffe: Deep learning for image classification
  •   Tensorflow: Open Source Software Library for Machine Intelligence (good software for deep learning)
  •   Theano: Deep learning library
  •   mxnet: Deep Learning library
  •   Torch: Scientific computing framework with wide support for machine learning algorithms
  •   LIBSVM: A Library for Support Vector Machines (Matlab, Python)
  •   scikit: Machine learning in Python

 

 

Courses

 

  •   Introduction to Neural Networks, CSC321 course at University of Toronto
  •   Course on Convolutional Neural Networks, CS231n course at Stanford University

 

 

Estimated learning time

Total number of hours 75

 

Face-to-face learning activities

30

 

-  Lecture with practical component

 

30

Supervised project

15

Independent learning

30

 

 

Competences to be gained during study

 

To get the basic knowledge and understanding that provides a basis or opportunity for originality in developing and / or applying ideas, often in a research context

 

To be able to work in a team.

 

To be able of defining an hypothesis from an insight by using exploratory analysis.

 

To be able of validating an hypothesis by hypothesis testing.

 

To be able to access bibliographic databases in order to find rlevant information.

 

 

 

 

Learning objectives

 

Referring to knowledge

Machine learning is one of the fastest-growing and most exciting fields out there, and deep learning represents its true bleeding edge. In this course, you’ll develop a clear understanding of the motivation for deep learning, and design intelligent systems that learn from complex and/or large-scale datasets.

 

 

Teaching blocks

 

1. Introduction

*  Software environments: Jupyter, Docker. Classification/regression. Validation Set. Score function. Loss function. Softmax. Optimization (SGD)

2. Basic Algorithms.

*  Automatic differentiation. Backpropagation. 

3. Tensorflow programming model.

4. Tensorflow ecosysgtem: Scikit Flow. TFLearn.

5. Convolutions. CNN models images.

6. Convolutions. CNN models for times series, and text.

7. Recurrent Neural Netwoks for text.

8. Recurrent Neural Netwoks for images.

9. Unsupervised Learning.

10. Reinforcement Learning

 

 

Teaching methods and general organization

 

All sessions will follow a theoretico-practical approach.

 

 

Official assessment of learning outcomes

 

Grading: Problem Sets: 70%. Projects: 30% (Oral Presentation: 5%, Project report: 25%) 

 

Examination-based assessment

Grading: Problem Sets: 100%.