PracticumAI / rnn

Practicum AI Recurrent Neural Network workshop.

Workshop Learning Objectives (Recurrent Neural Networks)

Session 1

1. Operationally define recurrent neural network (RNN).
2. Describe the purpose of recurrent neural networks.
3. Discuss where sequences, order, and remembering of samples/data.
4. Discuss time series data.
5. Describe Numpy.
6. Describe Pandas.
7. Describe Matplotlib.
8. Implement visualizing time-series data
9. Discuss the results and interesting outcomes or surprises.
10. Discuss the RNN process.
11. Describe state.
12. Implement your first plain RNN model.
13. Discuss the results and interesting outcomes or surprises.

Session 2

1. Discuss how context is important to language.
2. Describe the vanishing gradient problem.
3. Describe the exploding gradient problem.
4. Describe long short term memory (LSTM) architecture.
5. Discuss different RNN architectures.
6. Implement sentiment classification models.
7. Discuss the results and interesting outcomes or surprises.
8. Discuss and compare the five different CNN models.