This week we decided to pivot from our original theoretical direction to a different empirical investigation. I finished implementing the BIWZ and Random Rebucketing Extractor from Periklis’ paper from 9 years ago. Took some time relearning C++ and figuring out bit manipulation. I also found some quantum randomness sources, but there isn’t enough data to do anything meaningful. I will probably need to look into other sources of randomness.
The current line of investigation is to try to poison the randomness generated from this extractor, and see how it behaves when thrown into the NIST Randomness tests and a neural data compressor. Ultimately we want to see if we can develop a better test with this method.
Sources:
- True Randomness from Big Data (https://www.nature.com/articles/srep33740)
- CURBy Randomness Beacon (https://random.colorado.edu/)