Linearized sigmoidal activation: A novel activation function with tractable non-linear characteristics to boost representation capability
作者：
Highlights：
• An activation function with data-range-segment-wise nonlinear behavior is proposed.
• Values of slope coefficients for the proposed function are evaluated empirically.
• Proposed activation function is presented in adaptive and non-adaptive variants.
• Performance of proposed function is evaluated against state-of-the-art.
• Internal activation behavior and convergence characteristics are analyzed.
摘要
•An activation function with data-range-segment-wise nonlinear behavior is proposed.•Values of slope coefficients for the proposed function are evaluated empirically.•Proposed activation function is presented in adaptive and non-adaptive variants.•Performance of proposed function is evaluated against state-of-the-art.•Internal activation behavior and convergence characteristics are analyzed.
论文关键词：Deep learning,Activation function,Non-linear activation,Adaptive activation,Convolutional neural networks
论文评审过程：Received 5 September 2018, Revised 13 November 2018, Accepted 29 November 2018, Available online 30 November 2018, Version of Record 5 December 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.11.042
