ML Anthology

Reducing Neural Network Parameter Initialization into an SMT Problem (Student Abstract)

Mohamad H. Danesh
AAAI 2021 pp. 15775-15776
doi:10.1609/AAAI.V35I18.17884 /aaai/2021/danesh2021aaai-reducing/

Abstract

Training a neural network (NN) depends on multiple factors, including but not limited to the initial weights. In this paper, we focus on initializing deep NN parameters such that it performs better, comparing to random or zero initialization. We do this by reducing the process of initialization into an SMT solver. Previous works consider certain activation functions on small NNs, however the studied NN is a deep network with different activation functions. Our experiments show that the proposed approach for parameter initialization achieves better performance comparing to randomly initialized networks.

PDF AAAI Semantic Scholar

Cite

Text

Danesh. "Reducing Neural Network Parameter Initialization into an SMT Problem (Student Abstract)." AAAI Conference on Artificial Intelligence, 2021. doi:10.1609/AAAI.V35I18.17884

Markdown

[Danesh. "Reducing Neural Network Parameter Initialization into an SMT Problem (Student Abstract)." AAAI Conference on Artificial Intelligence, 2021.](https://mlanthology.org/aaai/2021/danesh2021aaai-reducing/) doi:10.1609/AAAI.V35I18.17884

BibTeX

@inproceedings{danesh2021aaai-reducing,
  title     = {{Reducing Neural Network Parameter Initialization into an SMT Problem (Student Abstract)}},
  author    = {Danesh, Mohamad H.},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2021},
  pages     = {15775-15776},
  doi       = {10.1609/AAAI.V35I18.17884},
  url       = {https://mlanthology.org/aaai/2021/danesh2021aaai-reducing/}
}