Skip to content
Conference Open access

Hybrid Machine Learning Approach for the Prediction of Esophageal Cancer using Random Forest and CNN

Sep 2026 · 2026 10th International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC) · pp. 857-862 · 0 citations · 10 references

Abstract

Esophageal cancer ranks as one of the deadliest cancers because it is usually found in its advanced stages, where there is no distinguishing feature indicating the presence of esophageal cancer in the initial phases. Therefore, early diagnosis and prediction are crucial to have better chances of survival. In this paper, we present an integrated machine learning algorithm integrating Convolutional Neural Networks (CNN) with a Random Forest (RF), aimed at predicting esophageal cancer through medical images and clinical features. Specifically, the CNN is employed to automatically remove features from the medical images, including endoscopic, identifying deep-spatially located features that can be signs of cancer. After extracting these features, the CNN passes them to RF in combination with other clinical features and makes a prediction on whether esophageal cancer exists. This paper leverages the unique benefits provided by each algorithm, i.e., deep feature extraction from CNN and efficient high-dimensional feature classification from RF. Our experimental results indicate enhanced predictive capability of the combined hybrid framework over standalone models, suggesting the use of CNN-RF for esophageal cancer prediction.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.