Research

On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,051 papers · 148 categories

Trend · papers per month

count % of corpus

1 result for “BDFA”

Efficient Convolutional Neural Network Training with Direct Feedback Alignmentcs.LG

Improved CNN training with BDFA reduces computational cost and improves accuracy.

problem Low training performance of DFA in CNN.

method Combining DFA with BP, introducing feedback weight initialization, and proposing BDFA.

result BDFA shows better performance than conventional BP, especially in small datasets.

CNN training BDFA backpropagation feedback alignment small dataset

2019-01-06abs ↗pdf ↗