Course Description
( upcoming/recent | all | 2024 | 2023 | 2022 | 2021 | 2020 | 2019 | 2018 | 2017 | 2016 | 2015 | 2014 | 2013 | 2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006)
Course on Deep Learning / Tensor Flow (16 hours)
Speaker:
Prof. Romain Herault (INSA, Rouen, France)
Dates:
- Monday, May 20, and Tuesday, May 21, 2019: 6 hours from 9h30 to 17h30 (lunch break, from 13h to 14h)
- Wednesday, May 22, 2019: 4 hours from 10h to 15h30 (lunch break, from 12h to 13h30)
Registration:
Free but mandatory.
Registration is now closed. The maximal number of possible attendees has been reached.Note that the number of seats is limited to 40.
Location:
"Shannon" Seminar Room, Place du Levant 3, Maxwell Building, 1st floor
Map: https://tinyurl.com/LocDLTFlow
Abstract and course content:
Neural Network and most notably Deep Learning are getting more and more popular outclassing ad-hoc states of the art methods in Image Processing, Natural Language Processing, Pattern Recognition… In this class, we propose to introduce you to this (not so new) kind of machine learning model.
Lectures (4*1h30) will address the following concepts:
- Introduction to Artificial Neural Networks and basic supervised
learning model- Unsupervised learning models such as Auto-Encoders
- Deep Neural Networks and Convolutional Neural Networks
- Recurrent Neural Networks (RNN)
- Generative Adversarial Networks (GAN)
Practical works (4*1h30) will implement the models seen in lectures through Tensorflow/Keras and provide analysis of the learning through Tensorboard/Spyder.
A longer project/practical work (2*2h) will linger on GAN or RNN.
Computing environment:
Persons attending this class are expected to come with their own laptop with Tensorflow/Keras/Tensorboard installed.
Pip packages needed:
spyder numpy scipy matplotlib ipython scikit-learn pandas scikit-image pillow pillow-pil tensorflow tensorflow-tensorboard keras cvxpy cvxopt pykalman pomegranate h5py spyder-terminal
For Linux and Mac OS users, you can follow this procedure:
Add this line at the end of your $HOME/.bashrc file
export PATH=$HOME/.local/bin:$PATH
Login/logout or open a new terminal, and type the following commands
$ wget https://bootstrap.pypa.io/get-pip.py
$ python3 get-pip.py --user
$ python3 -m pip install --user -U pip virtualenv pipsi wheel
$ pipsi install --python python3 spyder
$ source ~/.local/venvs/spyder/bin/activate
(spyder) $ pip install -U numpy scipy matplotlib ipython
(spyder) $ pip install -U numpy scipy matplotlib ipython scikit-learn pandas scikit-image pillow pillow-pil tensorflow tensorflow-tensorboard keras cvxpy cvxopt pykalman pomegranate h5py spyder-terminal
(spyder) $ pip install -U spyder
(spyder) $ deactivate
Not tested:
It is also possible to install these packages with the Anaconda environment