people:taylor_berger
Taylor Berger
Bullets are cleaner:
- First semester graduate
- Interests include creating robust software, machine learning, statistics and probability applied to computer science
- Github is here: www.github.com/teberger/
Paper: Link to paper's pdf on Google Drive. Project Materials: materials (Dropbox link)
Project Ideas:
- Low Fidelity Data Compression: Thinking in the MFM framework is hard. Through iterative
failuresproject evaluations, I realized I was trying to implement an algorithm in a robust first framework. Sorting in MFM is a great example of the issues with implementing algorithms – it doesn't make sense to recreate quicksort in the MFM, however Demon Horde Sort was created as a procedural sorting tool. A neural network is an algorithm we use – it doesn't make sense to recreate it in the MFM (or maybe it does, but I'm not sure how at this point). In light of that realization, I took what I learned and leveraged the work I have done so far to change my project to low-fidelity data compression. - The more realistic brain: After going through the first cycle of design in ALife, this is better scaled project due to the time constraints of the semester
- Biologically Inspired, Infinitely Scaleable Brain: Growing a biologically inspired, infinitely scaleabe neural network in the MFM with dendrites and axons and synapses, oh my!
- Adaptive immune system for the MFM based on biologically inspired immune systems. (Antigens and white blood cell style)
News
07-Dec-2014 09:00:18PM-0700
- Updated this page to include a new link to my finished paper
10-Nov-2014 12:55:06PM-0700
- It's been a while since this was updated. I am updating my project pages to document the failu… I mean, successful elimination of project topics.
07-Sep-2014 08:49:18PM-0600
- Added a page describing my preferred project
26-Aug-2014 01:18:41PM-0600
- Project Ideas updated
August 22, 2014
- I learned how to wiki and created this page
people/taylor_berger.txt · Last modified: 2014/12/09 00:37 by tberge01