I chose to develop a machine-learning model that could identify a cat from dogs in order to only allow cats in through a door. I had to upload collages of both cats and dogs. Each data set required at least 6 different collages in order for there to be enough data for the machine to work. The model was an image training model, so only images would have worked. Other data sets could have been useful such as pictures of both cats and dogs in one image, as well as data sets where there were other animals such as squirrels, mice, rabbits, etc. My model was decent for what the goal was. It was usually accurate in its predictions. The model did however struggle when trying to identify odd-looking cats/dogs. This problem could be easily fixed, however, this could require hundreds of thousands of images.