DeapSECURE module 4: Deep Learning (Neural Network): Instructor Notes

Lessons in Neural Networks & HPC

This lesson module serves several important purposes:

1. Introduce learners to the basic concepts of neural networks for classification tasks, and the basic procedure to construct, train, evaluate, and improve the neural network models.

2. Reinforce the initial data exploration, cleaning, and preprocessing with a new dataset, sherlock_18apps (which nevertheless bears a lot of similarities to the “2-apps” dataset).

3. Learn the process of converting an interactive Jupyter analysis to a noninteractive script.

This lesson would truly bridge people from the “interactive computing mode” on Jupyter platform to the “batch noninteractive computing” on HPC with job scheduler.