Enhancing Cancer Cell Classification through Dataset Augmentation using Conditional Variational Autoencoder (CVAE)

Anita Esi Eshun

Keywords: CVAE, cancer cell, augmented dataset

TL;DR: Achieving higher classification accuracy power through data augmentation

Abstract:

This paper uses Conditional Variational Autoencoder (CVAE) as a data
augmentation technique for cancer cell classification using four different classification algorithm; Support Vector Machine (SVM), XGBoost, Decision Tree, and
Logistic Regression. The main aim of the study is to enhance the accuracy of the
classification models by generating synthetic data using CVAE. The results obtain
shows that there is significant improvement when CVAE is used to generate additional data. The approach used indicates a promise in accurate cancer classification,
even in the absence of adequate dataset. This will help to improve cancer diagnosis
and treatment in the health industry

Submission Category: Machine learning algorithms

Submission Number: 80