<p>Colon cancer is a type of cancer that occurs in the large intestine, and it increases the mortality rate of people. Histological examination is conducted for precise colon cancer diagnosis. Deep learning-based colon cancer classification methods are recommended by researchers, but they are not flexible, and the time complexity of these approaches is high. Therefore, in this work, an efficient deep learning-based approach is developed to classify colon cancers into various types. First, the images required to perform colon cancer classification are obtained from the Kaggle source. Then, the collected images are directly fed to the developed Vision Transformer with Adaptive Neural Architecture Search (ViT-ANAS) to perform the colon cancer classification process. The developed ViT-ANAS approach leverages the strength of the Vision Transformer (ViT) and Neural Architecture Search (NAS) for the effective classification process. Here, ViT extracts the features from the images, and then the extracted features are given to the NAS to get the classified outcome. The efficient neural structure in Neural Architecture Search (NAS) is optimally selected using the Enhanced Golden Jackal Optimization (EGJO) to perform the effective colon cancer classification. In addition, the parameters such as learning rate and epoch size are tuned using the EGJO to improve the accuracy of colon cancer classification. Final classification results are compared with the conventional colon cancer classification approaches to ensure effectiveness.</p>