We Ran 35 Controlled Experiments on "Exotic Algebra" Neural Networks. They're a Placebo.

This story was originally published on HackerNoon at: https://hackernoon.com/we-ran-35-controlled-experiments-on-exotic-algebra-neural-networks-theyre-a-placebo. Quaternion, octonion, sedenion, and Clifford networks lose their edge when tested against properly matched real-valued baselines. Check more stories related to machine-learning at: https://hackernoon.com/c/machine-learning. You can also check exclusive content about #machine-learning, #neural-networks, #ai-research, #negative-results, #mathematics, #hypercomplex-networks, #quaternion-neural-nets, #hackernoon-top-story, and more. This story was written by: @exactor. Learn more about this writer by checking @exactor's about page, and for more stories, please visit hackernoon.com. Quaternion, octonion, sedenion, and Clifford networks lose their edge when tested against properly matched real-valued baselines.