Skip to content

Home

Jump to bottom
osu-msban-ai edited this page Sep 7, 2018 · 16 revisions

Welcome to the Deep-Learning-Course-Files wiki! Please use the sidebar on the right to navigate.

Main Page

The objective of this short course is to equip students with basic concepts and tools to work on real-world problems. This course is not designed to provide in-depth mathematical coverage of the concepts and neither is it a coding boot camp. In this course, we will use Tensorflow from Google as the basic deep learning framework upon which to build our models. In addition, we will use Keras, which is a higher level abstraction on top of Tensorflow. In practical situations, Keras might be a more efficient interface to use when building complex deep learning models. Except for the introductory session, all subsequent work will be done in Keras.

The prerequisites for this course will be:

  1. Coding capability in Python
  2. Exposure to Numpy, a python linear algebra library
  3. Pandas – the data management library on python
  4. Basic concepts of machine learning. Online resources are available to catch up on these, if necessary.

The coding platform we will use is the Jupyter Notebook. For local use, the infrastructure will be a preset docker container containing all the necessary libraries. For models which will require GPU based computing, we will use Amazon's AWS cloud platform, especially the p2 GPU instance or similar local resources available from OSU. This wiki provides a list of resources and 'how-tos' for the course.

  1. Course Outline

    1.1 Part 1 - Deep Neural Network

    1.2 Part 2 - Convolutional Neural Network

    1.3 Part 3 - Recurrent Neural Network

    1.4 Part 4 - Project

  2. How-To's

    2.1 Set up Docker container on Local Machine

    2.2 Setup AWS instances

  3. Resources

    3.1 Additional Material for CNN Demo

    3.2 Libraries and Frameworks used for the Course

    3.3 Other Useful Resources

Clone this wiki locally