AN EXTENDED FRAMEWORK OF LUNG CANCER CLASSIFICATION USING HYBRID ARCHITECTURE OF SURF AND SVM

Main Article Content

Paramjit Singh, et. al.

Abstract

The present research work focussed on the Lung Cancer disease classification by the potential usage of hybrid model in which segmentation, feature extraction, optimization and classification techniques has been performed on the dataset of CT scan images of 1000 images. A set of 1000 images are to be utilized in which 75% data is used for the test purpose and rest 25% is used for classification. The present research article measured the performance of hybrid model by applying the post segmentation techniques Particle Swarm Optimization (PSO),Artificial Bee Colony(ABC), FFA(Fire Fly Algorithm) Cuckoo Search(CS), and best features extraction technique Speed Up Robust Feature (Surf) in the terms of minimum execution time and minimum error rate with classifier Support Vector Machine(SVM) is also used as cross validator for the evaluation of the performance of hybrid model in the terms of parameter accuracy, error rate , precision, recall and execution time. The overall accuracy of hybrid model has 98.90%, recall value 91.46%, F-measure 94.73% and minimum execution time .00031 secs has been achieved for the hybrid model.

Article Details

Section
Articles