Scale-invariant learning and convolutional networks

详细信息    查看全文

• 作者：Soumith Chintala ^{soumith@fb.com} ; Marc'Aurelio Ranzato ^{ranzato@fb.com} ; Arthur Szlam ^{aszlam@fb.com} ; Yuandong Tian ^{yuandong@fb.com} ; Mark Tygert ; ^{tygert@fb.com} ; Wojciech Zaremba ^{wojciech@fb.com}
• 关键词：Spectrum ; Wavelets ; Wavelet packets ; Signal processing ; Image processing ; Classification ; Equivariant ; Invariant ; Machine learning ; Representation
• 刊名：Applied and Computational Harmonic Analysis
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：42
• 期：1
• 页码：154-166
• 全文大小：1565 K
• 卷排序：42

文摘

Multinomial logistic regression and other classification schemes used in conjunction with convolutional networks (convnets) were designed largely before the rise of the now standard coupling with convnets, stochastic gradient descent, and backpropagation. In the specific application to supervised learning for convnets, a simple scale-invariant classification stage is more robust than multinomial logistic regression, appears to result in somewhat lower errors on several standard test sets, has similar computational costs, and features precise control over the actual rate of learning. “Scale-invariant” means that multiplying the input values by any nonzero real number leaves the output unchanged.