scholarly journals MACHINE LEARNING TECHNIQUE TO PREDICT THE MODEL FOR NON-SMALL CELL LUNG CANCER-BINARY LOGISTIC REGRESSION MODEL APPROACH

2021 ◽  
pp. 28-29
Author(s):  
K.Vaishnavi Devi ◽  
S. Venkatesan

Lung cancer is one of the common types and deadly cancer in both men and women.This lung cancer accounts for high mortality and morbidity throughout the world.Detection of lung cancer has been made through surgery,chemotherapy, biopsy and microarray studies.Gene expression plays an important role in molecular fluctuations and disease prophecy of a disease.The aim of the study is to design a statistical model and to find the genes influencing the cause of lung cancer. Microarray gene expression data was collected from Gene Expression Omnibus datasets (GEO-DATASET)-an open source database.The dataset contains a total of 161 samples which has 89 lung cancer samples and 72 normal samples. From this the upregulated and influenced genes were identified and determined by using logfc from the GPL file.Wide use of statistical models leads to exploring machine learning methods to find a better model. These study methods implement the performance of regression analysis using multilayer perceptron. By using the regression analysis method,the overall accuracy is found to be 91.3%.By this,the gene expression data analysis reveals that the regression analysis is one of the best models to show the accuracy in implementation of genes influencing the NSCLC.

Sign in / Sign up

Export Citation Format

Share Document