You’re starting a new venture firm MACHINES ARE SMART, that specializes in machine learning. You are going to pitch a new project to venture capital investors, and you need a proof of concept.  Use teachable machine to build a model that does one of the following:

• A special pet door that only lets in only cats, for cat lovers.
• A foreign language training app that teaches you to say “hello” in a foreign language and rejects attempts that aren’t correct. 
• An app that rejects drinks that aren’t coffee (We chose this option).
• An app that tells if elderly nursing home residents are happy or sad, and alerts loved ones if the nursing home resident is too sad too often.

Reflect on the following question: You’ll have to think about how to train the machine.  What kind of data did you include in your training dataset and why? What other kind of data could have been helpful but maybe you couldn’t get in the short-term/for free? Your group may, in some cases, search for photograph sets.  One possibility to get large data sets is to convert YouTubes into clips. Did your model work well for what you wanted?  In what instances might your model not work very well?  Include the link to your project.

For our dataset, we included data of images of coffee in varying cups, with milk, without will, hot, iced, with straws, from Starbucks, and from Dunkin. We also included data of images other than coffee such as juice, smoothies, wine, beer, lemonade, and water. It would have been helpful to get these substances in more cups and from more angles, but short-term this was not possible. We searched google for images and chose a variety of pictures, instead of YouTube videos. This worked well for us and our teachable machine is pretty accurate, but there weren’t many substances other than water in the room to test the model on. Our model may not work well in cups that are opaque or tinted and change the color of the drink.