Dear students,
I hope you have been enjoying the Artificial Neural Networks course so far! As we approach the middle of the semester, it is the time of the year that we release the mini-projects. You should submit ONE of the two mini-projects in a group of TWO students. This will make 30% of your final grade. See below for short descriptions of the mini-projects (and see Moodle for how to set up the Keras environment):
In mini-project1: You will be using REINFORCE algorithm to train a neural network on an Atari game (Lunar Lander). For this project, you can choose either to do the project on your computer or to do it over Google Colab.
In mini-project2: You will be training feedforward and convolutional neural networks on standard and noisy datasets. You will be practicing the mechanisms and tricks that you have started learning in week 5.
We encourage you to use the related Slack channels for the mini-projects to exchange ideas and ask your questions to the TAs.
Have a nice evening,
Berfin