KTBYTE Computer Science Academy >> Classes >> Advanced >> Cs84 Deep Learning
NOTICE ISP is back up.. Stay tuned for class status for Thursday 2017-03-23

CS84: Deep Learning

Prerequisites: Completion of CS01b or AP CS, or permission of instructor. Also requires Algebra II math experience. CS82 highly recommended but not required.

Learn the most modern techniques for supervised learning, used in common applications such as facial recognition, speech recognition, and self driving cars. This course will also provide students with a linux server with GPU acceleration to run their algorithms. Topics include test classification, convolutional image recognition, q-learning, and more.

This course uses the Keras deep learning library.

Course Content

  1. (Skipped this semester) Review of Python, Numpy, and Jupyter Notebook
  2. Using Linux
    • Recommended Reading from Cornell: lec02 lec03
    • Linux Command Line Tutorial, WinSCP, XMing
    • (Last semester) Homework: Write a program to conditionally average a column in a CSV in at least 5 different languages: Java, Python, Bash, R, and awk
    • (This semester) Homework: 2017-02-7 Read historical machine learning papers, pick favorites
  3. Using Keras and python
  4. Using Keras for Classification Part 2
  5. Convolutional Networks, MNIST
    • Homework Pt 1 (Start in class): Kaggle Contest
    • Homework Pt 2 Shape Count
    • (Alternative homework) Shape Points. Continue the in-class project "Shape Points". Modify the shapePoints project to draw two circles and several squares. Output an image representing the distance from the closest circle for each pixel (in manhattan distance, not euclidian). Build a neural network to predict the manhattan distance. Write a processing program that loads the test data and predictions data, and draws 4 pictures side by side. Picture 1 should be the input. Picture 2 is the correct output. Picture 3 is the predicted output. Picture 4 is the difference between pictures 2 and 3.
  6. Captcha Breaking (chars74k), Image Pre-Processing
  7. Cats versus Dogs
  8. Cats versus Dogs pt 2
  9. The Titantic Data Set
  10. Word Vectorization and gensim, Text Classification, GloVe
    • Demo: News Groups
    • Homework: Parse IMDB Sentiment Data into files/folders
  11. Text Parsing, IMDB Sentiment Analysis
    • Homework: Finish IMDB
  12. Q Learning Part 1
  13. Q Learning Part 2 - Games
  14. Paper and Project Review
  15. Paper and Project Discussion (Favoriate Topics)
  16. Paper and Project Discussion (Favoriate Datasets)
  17. Independent Project: Data Munging
  18. Independent Project: Data Munging
Format: Web-conference

See Class Times