This is Part 3 of a four-part series that breaks up a talk that Seth Juarez gave at the Toronto AI Meetup.

Parts 1 and 2  introduce basic machine learning concepts as well as specific models using TensorFlow respectively.

In this video he goes more in depth into an example of a common data science process, how convolutions work in convolutional neural networks, and finally how this can be done in the cloud using Azure Machine Learning.

Siraj Raval explores why does a computer algorithm classify an image the way that it does? This is a question that is critical when it comes to AI applied to diagnostics, driving, or any other form of critical decision making.

In this video, he raises awareness around one technique in particular that I found called “Grad-Cam” or Gradient Class Activation Mappings.

Siraj Raval shows off examples of machine learning apps from his students.

If you’re wondering about my stance on the recent controversies around Siraj, I recorded a Data Point about that.

Machine Learning powers almost every internet service we use these days, but it’s rare to find a full pipeline example of machine learning being deployed in a web app. In this episode, I’d like to present 5 full-stack machine learning demos submitted as midterm projects from the students of my current course. The midterm assignment was to create a paid machine learning web app, and after receiving countless incredible submissions, I’ve decided to share my favorite 5 publicly. I was surprised by how many students in the course had never coded before and to see them building a full-stack web app in a few weeks was a very fulfilling experience. Use these examples as a template to help you ideate on potential business ideas to make a positive impact in the world using machine learning. And if you’d like, be sure to reach out and support each of the students I’ve demoed here today in any way can you offer. They’ve been working their butts off. Enjoy!

Presentation notebook: https://colab.research.google.com/drive/1m5aLHPnwIhVX8zgMvZUtK4iG9xSaMbk8