Application of Chicken Swarm Optimization in Detection of Cancer and Virtual Reality

Abstract

Cancer is a very common type of disease occurring amongst people and it is also amongst the main causes of deaths of humans around the world. Symptom awareness and needs of screening are very essential these days in order to reduce its risks. Several machine learning models have already been proposed in order to predict whether cancer is malignant or benign. In this paper, we have attempted to propose a better way to do the same. Here we discuss in detail about how we have applied the chicken swarm Optimisation as a feature selection algorithm to the cancer dataset of features in order to predict if the cancer is malignant or benign. Here we also elucidate how the Chicken Swarm Optimization provides better results than several other machine learning models such as Random Forest, k-NN, Decision Trees and Support Vector Machines. Feature Selection is a technique used to eliminate the redundant features from a large dataset in order to obtain a better subset of features to use for processing. In order to achieve this, we have used Chicken Swarm Optimization. The chicken swarm optimization algorithm is a bio-inspired algorithm. It attempts to mimic the order of hierarchy and the behavior of chicken swarm in order to optimize the problems. On the basis of these predictions we can also provide quick treatment by using virtual reality simulators that can be helpful for complex oncological surgeries. The results shown by this are better than the other models as this model achieves a very high accuracy as compared to the others discussed in the paper.