Accurate and timely diagnosis of colon diseases is crucial for effective treatment and better patient outcomes. Wireless Capsule Endoscopy (WCE) offers a non-invasive method to visualize the gastrointestinal tract, including the colon. However, manually analyzing WCE images is time-intensive and prone to errors. This study presents a novel machine learning-based approach for diagnosing colon diseases using WCE images. The proposed method extracts statistical features from the images, such as texture, color histograms, and shape descriptors, to capture important visual characteristics linked to pathological conditions. These features are then used to train and validate machine learning classifiers that can distinguish between healthy and diseased tissues. The study assesses the performance of Ensemble Bagging and Multi-Kernel Support Vector Machine learning model, in disease prediction. Experimental results demonstrate that the proposed approach achieves high accuracy, sensitivity, and specificity, underscoring its potential as an effective tool for automated diagnosis. By combining machine learning with WCE image analysis, this approach could enhance diagnostic accuracy, alleviate the workload of healthcare professionals, and ultimately improve patient care.