TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK	EXTRA DATA	REMOVE
Image Classification	CIFAR-10	NiN+APL	Percentage correct	92.5	# 162
Image Classification	CIFAR-100	NiN+APL	Percentage correct	69.2	# 166

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-activation-functions-to-improve-deep/image-classification-on-cifar-10)](https://paperswithcode.com/sota/image-classification-on-cifar-10?p=learning-activation-functions-to-improve-deep)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-activation-functions-to-improve-deep/image-classification-on-cifar-100)](https://paperswithcode.com/sota/image-classification-on-cifar-100?p=learning-activation-functions-to-improve-deep)`

Learning Activation Functions to Improve Deep Neural Networks

21 Dec 2014 · Forest Agostinelli, Matthew Hoffman, Peter Sadowski, Pierre Baldi ·

Artificial neural networks typically have a fixed, non-linear activation function at each neuron. We have designed a novel form of piecewise linear activation function that is learned independently for each neuron using gradient descent. With this adaptive activation function, we are able to improve upon deep neural network architectures composed of static rectified linear units, achieving state-of-the-art performance on CIFAR-10 (7.51%), CIFAR-100 (30.83%), and a benchmark from high-energy physics involving Higgs boson decay modes.

PDF Abstract