This workshop will cover material ranging from what a vector is all the way to Lp norms, loss functions, and gradient descent! We want to emphasize that a strong math background is not required for this workshop, as we'll be presenting the material in a beginner oriented, hands-on way. That means that we will introduce material both in terms of what you may code up in any given project, and the abstract math objects which represent them. In the simplest case, a vector can be described as a 1D array, but that's not enough to justify many of the techniques employed in DL. In order to extend that, we will dive into the math that powers the code.
Please fill the sign-up sheet below https://forms.gle/DV2dQ7ghQSvYWsEG7
- While in your command line, move to a directory that you want to clone the workshop into.
- Simply type
git clone https://github.com/delug/Workshop2.gitin your command line to clone the repository - Run
jupyter notebookand navigate to where you cloned the workshop repository - Open the notebook and enjoy!
Note: Before the workshop, please make sure you have the most up-to-date version of this repository. This can be assured by running git pull
within the repository close to the workshop day. Preferably the day of, just to be safe!
Before coming to the workshop, please ensure that you have the following softwares downloaded:
- Python (We recommend downloading Python along with Anaconda: https://www.anaconda.com/distribution/)
- Jupyter (https://jupyter.org/install)
- Numpy (In command line, enter:
pip install --user numpy) - Git (https://git-scm.com/downloads)
Deep Learning at UGA is a club that began as a small organization and is rapidly expanding to service as many people as possible. This is a difficult task, as we're often breaking new ground and sometimes it shows. We want to ensure that everything we offer is of the highest possible quality, but that requires help from you! If you've got a spare second, it would mean a lot if you could take the survey below to share your feedback with us. We go through every single response and work to meet your needs. Please fill out the survey in the link below!