:This paper presents the hybridization of Whale
Optimization Algorithm (WOA) with Extreme Learning
Machine (ELM) methodology for solving Gene classification
problem. ELM is assumed to be a likely technique for prediction
and classification problems. Despite its effectiveness, it needs a
large number of nodes on a regular basis for the hidden layer.
Using such a huge number of nodes within the hidden layer
increases the ELM examination/assessment time. In addition,
there is a little guarantee that the layout of weights and biases
inside the hidden layer would be optimum. A recent swarm
intelligence algorithm (WOA) mimics the conduct of the hunting
party of humpback whales is proposed to optimize the ELM
model. It is being used within the hidden layer to pick a smaller
number of nodes to accelerate the execution of ELM. WOA
chooses the optimal weights and bias of the hidden layer.
Experimental results show that the proposed hybrid model
(WOA-ELM) had better classification accuracy than the
standard ELM and SVM. |