The Kramer-Eden-Chu lab hackathon to use deep learning to create an automated spike-ripple detector. This builds on the project in Nadalin et al., 2021. Exact aims will depend on what we decide to do during the hackathon, but some examples include
- Use a model pre-trained on a different set of images (e.g., satellite imagery instead of natural images)
- Account for expected artifacts (e.g., broadband increases in power)
- Use an LSTM-based RNN
- Augment the training dataset
-
Nadalin, Jessica K., Uri T. Eden, Xue Han, R. Mark Richardson, Catherine J. Chu, and Mark A. Kramer. 2021. “Application of a Convolutional Neural Network for Fully-Automated Detection of Spike Ripples in the Scalp Electroencephalogram.” Journal of Neuroscience Methods 360 (August): 109239. https://doi.org/10.1016/j.jneumeth.2021.109239. [Code]
-
Hagen, Espen, Anna R. Chambers, Gaute T. Einevoll, Klas H. Pettersen, Rune Enger, and Alexander J. Stasik. 2021. “RippleNet: A Recurrent Neural Network for Sharp Wave Ripple (SPW-R) Detection.” Neuroinformatics 19 (3): 493–514. https://doi.org/10.1007/s12021-020-09496-2. [Code]
-
Sarmashghi, Mehrad, Shantanu P. Jadhav, and Uri T. Eden. 2022. “Integrating Statistical and Machine Learning Approaches for Neural Classification.” IEEE Access 10: 119106–18. https://doi.org/10.1109/ACCESS.2022.3221436.
-
Free deep learning ebook
-
Fastai (course, tutorials, fastai for PyTorch)
- Coincidentally, this interview with Jeremy Howard of fastai just came out - he's very interested in making deep learning and AI accessible
.gitignore
Globally ignored files bygitfor the project.environment.yml
condaenvironment description needed to run this project.README.md
Description of the project (see suggested headings below)
Each team member has it's own folder under contributors, where they can work on their contribution. Having a dedicated folder for each person helps to prevent conflicts when merging with the main branch.
Notebooks that are considered delivered results for the project should go in here.
Helper utilities that are shared with the team
Brief title describing the proposed work.
List all participants on the project. Choose one team member to act as project lead, and identify one hackweek organizer as the data science lead.
What problem are you going to explore? Provide a few sentences. If this is a technical exploration of software or data science methods, explain why this work is important in a broader context.
List one specific application of this work.
If you already have some data to explore, briefly describe it here (size, format, how to access).
List the specific tasks you want to accomplish or research questions you want to answer.
How would you or others traditionally try to address this problem?
Building from what you learn at this hackweek, what new approaches would you like to try to implement?
- Nadalin, Jessica K., Uri T. Eden, Xue Han, R. Mark Richardson, Catherine J. Chu, and Mark A. Kramer. 2021. “Application of a Convolutional Neural Network for Fully-Automated Detection of Spike Ripples in the Scalp Electroencephalogram.” Journal of Neuroscience Methods 360 (August): 109239. https://doi.org/10.1016/j.jneumeth.2021.109239. [Code]