arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2301.07285cs.LG
收藏

一种新型稀疏正则化器

A Novel Sparse Regularizer

Hovig Tigran Bayandorian

AI总结:

本文提出一种基于新型熵度量、关注权重空间排列的稀疏正则化器,可在 MNIST 上训练 LeNet300 时将达到给定准确率所需的非零参数减少约一个数量级。

摘要:

诸如 $L_0$、$L_1$ 和 $L_2$ 范数正则化之类的 $L_p$ 范数正则化方案,以及基于 $L_p$ 范数的正则化技术(如权重衰减、LASSO 和弹性网),所计算的量依赖于彼此孤立考虑的模型权重。本文介绍了一种正则化器,其基础是在优化过程中对模型应用并最小化一种新型熵度量。与基于 $L_p$ 范数的正则化不同,该正则化器关注权重矩阵内权重的空间排列。这种新型正则化器是损失函数的一个附加项,具有可微性、计算简单快速、尺度不变性,只需极少的额外内存,并且可以轻松并行化。实验表明,在 MNIST 上训练 LeNet300 时,要达到给定测试准确率水平,该方法所需的非零模型参数数量约减少了一个数量级。

英文摘要:

$L_p$-norm regularization schemes such as $L_0$, $L_1$, and $L_2$-norm regularization and $L_p$-norm-based regularization techniques such as weight decay, LASSO, and elastic net compute a quantity which depends on model weights considered in isolation from one another. This paper introduces a regularizer based on minimizing a novel measure of entropy applied to the model during optimization. In contrast with $L_p$-norm-based regularization, this regularizer is concerned with the spatial arrangement of weights within a weight matrix. This novel regularizer is an additive term for the loss function and is differentiable, simple and fast to compute, scale-invariant, requires a trivial amount of additional memory, and can easily be parallelized. Empirically this method yields approximately a one order-of-magnitude improvement in the number of nonzero model parameters required to achieve a given level of test accuracy when training LeNet300 on MNIST.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑