Cervical Cancer Prediction and Classification using deep learning

EasyChair Preprint 2852

Abstract

Cervical cancer is one of the increasing sicknesses among women in India and also  around the world. Early analysis is good for better treatment, yet due to vulnerability in detecting cancer cells becoming more complex one. Machine Learning (ML) systems  were used to predict the cancer cells in human beings. For this method the cervical cancer  datasets  were taken from Unique Client Identifier (UCI) store  to predict the cancer cells . But this approach failed to provide better accuracy. In this paper, we propose a cervical cancer cell prediction and classification system based on deep learning techniques . Convolutional neural network (CNN) model is used for prediction and classification.To extract deep-learned features, the cell images were fed into a CNNs model. Further, the  input images were classified using an extreme learning machine (ELM)-based classifier. CNNs model is uses the methods namely, transfer learning and fine tuning for providing better accuracy. The experiment was done by collecting the cervical cancer dataset from pap smear Herlev database. Alternatives to the ELM, multi-layer perceptron (MLP) and autoencoder (AE)-based classifiers are also examined. The proposed CNN-ELM-based system achieved high accuracy in the prediction problem (2-class) and classification problem (4-class).

Keyphrases: Auto-encoder, Convolutional Neural Network, Extreme learning machne

@booklet{EasyChair:2852,
  author    = {Abdul Affiya and Gopalakrishnan Aiswarya and Gopalraj Brindha and S. Selvaraj},
  title     = {Cervical Cancer Prediction and Classification using deep learning},
  howpublished = {EasyChair Preprint 2852},
  year      = {EasyChair, 2020}}