Research Area:  Machine Learning

Abstract:

Deep learning is becoming an important method of machine learning and has proven to provide its strength in data sciences and artificial intelligence. It works with unstructured and unlabeled data. Deep learning is more accurate than conventional classifier-based machine learning. The reason for high accuracy is managing by a count of hidden layers, application of activation functions, weighted activation functions, etc. In this work, a new weighted activation function is being proposed to get high accuracy for deep learning. The weighted activation function is going to be applied and tested on a depression dataset to evaluate the intensity of depression.

Keywords:  

Author(s) Name:   Hrithik Sanyal; Sagar Shukla; Rajneesh Agrawal

Publisher name:  IEEE

DOI:  10.1109/ICCE50685.2021.9427624

Volume Information:  

Paper Link:   https://ieeexplore.ieee.org/abstract/document/9427624