Globally, gastric cancer is a major cause of cancer-related deaths, which highlights the importance of early detection in order to improve patient lifespan. The present preferred method for detection in clinical practice is histological image processing. However, this procedure is performed manually, requires significant effort, and is time-consuming. Consequently, there has been a rise in interest in the creation of pathologists-assisted computer-aided diagnostic systems. CNN, LR, and SVC have shown potential in this aspect, but each model is capable of extracting only a restricted amount of picture characteristics for categorization. In our research, we try to compare the models. In the world, there is a lot of research on this type of data. However, those research results are based on CNN model implementation. In our research, we try to implement other models, make comparisons, and find the best accuracy. The findings of our research demonstrated that the ensemble model, consisting of 5 models, attained the best level of detection success across all sub-databases, reaching an impressive accuracy of 96.33%, which is considered state of the art.