Schedule
Week 1
Wednesday – 17 January
- lecture: an Overview of AI and deep learning
- lecture: structure of class
- readings (these are from the popular press)
- The AI Revolution: The Road to Superintelligence – Tim Urban
- The AI Revolution: Our Immortality or Extinction – Tim Urban
- The Great AI Awakening – Gideon Lewis-Kraus, NYTimes (about Deep Learning)
- Inside The Mind That Built Google Brain: On Life, Creativity, And Failure – Nico Pitney, HuffPo
Friday – 19 January
- Today’s sole goal: to download and install the necessary software to run the digit recognition software.
- lab 1:
- download and Install Anaconda
- test that ipython works (type ipython and see if you get the python shell)
- install TensorFlow (conda install -c conda-forge tensorflow)
- clone the code repository https://github.com/martin-gorner/tensorflow-mnist-tutorial.git
- change into that directory tensorflow-mnist-tutorial
- execute your first tensorflow program ipython mnist_1.0_softmax.py
- celebrate
- help others
- sympathetic joy
- problem with tensorflow version mismatch? try this
- lecture: Introduction to Anaconda
- lecture: Introduction to Python
- Helpful Tutorial if unfamiliar with Python: Introduction to Python for Data Science
Week 2
Monday – 22 – January
- lecture: an insanely cool algorithm
- Blondie24: Playing at the Edge of AI (amazon) (public library)
- lecture: Introduction to Machine Learning
Wednesday – 24 January
- lecture: Genetic Algorithms (pdf of Slides)
Friday – 26 January
- lecture: Entropy . (pdf of my rough notes)
- lecture: Matrix Multiplication (pdf of notes)
- lab: 8 Bullet Friday Tensorflow (direct link to file)
- Sample Numpy Notebook if you need it (raw file)
Week 3
Monday – 29 – January
- continue on the 8 bullet Friday lab
- hands-on Tensorflow graphs and Tensorboard
- if time: Gradient Descent, Delta Rule, and Learning Rates
Wednesday – 31 January
- lecture: Activation Funcrtions (PDF)
- lecture: Gradient Descent
- lecture: Delta Rule
- lab: A complete Tensorflow program:
Friday – 2 February
- Resources
- But What is a Neural Network (19 minute video)
- Neural Networks (especially §10.2) – Alexandra Shtabnaya writes: I was surfing the internet looking for a good explanation of the delta rule and I found a great website that really helped clear my confusion. Sections 10.2 to 10.4 were particularly useful, and the website also provided some code examples for implementing the rule (albeit in Java). If anyone is still confused, I would recommend going to this article
- Lab
- Tensorboard error? Try:
- pip install –upgrade -I setuptools
- pip install –ignore-installed –upgrade tensorflow
- Tensorboard error? Try:
Week 4
Monday – 5 – February
- review of single layer neural networks.
- Take-A-Problem (pdf of questions)
Wednesday – 7 February
- Review continued
- Test Prep
Friday – 9 February
- test #1
Week 5
Monday – 12 – February
- Introduction to Keras
- multilayer networks
- youtube video explaining the meaning of layers
- Keras Python Notebook
- don’t forget to conda install keras
Wednesday – 14 February
- Deep Visual Semantic Alignments for Generating Image Descriptions
- Keras Python Notebook for Movie Reviews
Friday – 16 February
Week 6
Monday – 19 – February
Wednesday -21 February
- evaluation – False Positive, etc.
Friday – 23 February
- Dog and Cat Notebook – optional
Week 7
Monday – 26 – February
- Introduction to Convolutional Neural Networks – aka covnets & CNN (slides)
- Intro to CovNet Python Notebook (no xp just for reference)
Wednesday – 28 February
- covnets continued –
- lab – covnet with dogs and cats 150xp
Friday – 2 March
- class optional – open lab
Week 8
Monday – 5 – March
- spring break
Wednesday – 7 March
- spring break
Friday – 9 March
- spring break
Week 9
Monday – 12 – March
- Word Embeddings
- Word Embeddings Notebook
- 1D convolutional networks for text.
Wednesday – 14 March
- Understanding Recurrent Networks (slides)
Friday – 16 March
- Lab day
Week 10
Monday – 19 – March
- no class
Wednesday – 21 March
Friday – 23 March
- LSTM (slides)
Week 11
Monday – 26 – March
Wednesday – 28 March
Friday – 30 March
Week 12
Monday – 2 April
- Running on a GPU using Colab – notebook
- AutoML-CodeLab – this is pretty fun
- 75 xp
- 50 xp to train it on something other than clouds
Wednesday – 4 April
- Using a pre-existing CNN to recognize images: VGG16 notebook
Friday – 6 April
- Test Prep: Team Study Guide
Week 13
Monday – 9 April
Wednesday – 11 April
- Test – DNN, CNN, & RNN
Friday – 13 April
Week 14
Monday – 16 April
Wednesday – 18 April
Friday – 20 April
Week 15
Monday – 23 April
- Reinforcement Learning
Wednesday – 25 April
Friday – 27 April