GEPAF: A non-monotonic generalized activation function in neural network for improving prediction with diverse data distributions characteristics

Khush Attarde, Javed Sayyad

Abstract: Highlights•Proposed GEPAF offering a versatile solution for various data distribution characteristics.•Derivation and analysis of GEPAF properties with graphical representations.•Practical application of the GEPAF within an LSTM network model.•Utilizing real-life supply chain datasets for sales and profit prediction.•A comparative performance analysis of GEPAF and conventional functions was performed.