The Challenge It can be difficult shopping for a home for many reasons. One reason in particular is that a potential buyer might have a very specific style of home that they are interested in. House listing websites like Zillow classify their homes based on many features but don't include style.

The Solution It might be helpful to add a function that filters homes based on the exterior style of the home. This would be done with a computer vision classification model that classifies homes into a few different categories.

Results The model was trained on the Mobile Net v2 classification model and had 94% accuracy on the validation data.

Constraints The data is grouped into three types; farmhouse, modern, and rustic. There are many more different types of house types but to simplify things I just chose three styles that are common yet distinct from each other. Training on this dataset doesn't account for the fact that many homes are a combination of a few styles. Also I scraped the data from Google Images which had fewer modern and rustic homes than farmhouse homes