Representation Learning
In this assignment, you’re going to setup and familiarize yourself with the course work environment.
Obtain an account for this server from Siegfried Wrobel (room II.14.225, phone 0331/977-2331). You will have to visit him in person to set your password. You can find information on how to access the server, install additional software and run experiments in the course wiki (next step).
Visit the Campus.UP website and log in with your Potsdam University (UP) account. This will take you to your Dashboard page. Open the “workspaces” drawer on the left and search for “Representation Learning”. Join the workspace. This will send a request that needs to be approved. In case of problems, send an email to get an invitation to the workspace. Once you can access the course workspace, it will show up in your Campus.UP dashboard.
Next, access your profile through the menu at the top right. Optionally add a photo and some description about yourself.
Once you have access to the course workspace, have a look around. An individual blog will be created for you. Visit your blog and write a first post about yourself summarizing your experience with (deep) machine learning so far. For instance, you could also include links to your projects on Github.
As follow-up to the in-class discussion on course learning goals, please reflect on what your personal goals for the course are. Publish your goals as a dedicated post in your blog (e.g. as a bullet-point list).
Within the Campus.UP course workspace, access the course wiki. Read the information on the GPU compute server and follow the steps to access the jupyterhub. Any issues encountered should be posted in the course forum.
Starting from the 2nd assignment, fundamental working knowledge of the following tools will be required. Please use the 1st week to catch up where necessary. Use the course forum for help! Go at least through the basic examples and run them in jupyter.
General:
High-Level Frameworks: Pick at least one! If you have worked with one already, have a look at the other one.
High-Level Frameworks: Pick at least one! If you have worked with one already, have a look at the other one.
More resources will be posted in the course workspace.
You may also want to set up your own development environment on your own computer. This can be helpful to develop and test your code before you run it on medusa. Furthermore, you can also use this environment when you have no internet connection.
For next week’s class, please read Deep Learning Book - Chapter 6: Deep Feedforward Networks. Refer to Deep Learning Book - Part I: Applied Math and Machine Learning Basics for any points that need a refresh.
Post your remarks, insights and open questions in your blog!
Deadline for questions to be considered in class is October 24, 7am.