Inspiration

I was inspired to make this skill while taking a machine learning class here at college. While we are covering the basics of scikit-learn, we haven't really been focusing much on how to choose which algorithm and when to use it. So, when my professor presented a useful flowchart on how to choose (included in the image gallery), I knew I could feed its information to Alexa for a positive and time-saving learning experience.

What it does

It makes the most sense if you already have a project in mind. Your data set is chosen, and you have decided to classify, quantify, or examine your data. You don't need these prerequisites to use the skill, however. The skill walks you through a series of yes or no questions terminating in a recommendation for which machine learning algorithm would be useful in your situation.

How I built it

I built it entirely in Storyline, my most beloved online tool. This consisted of mainly building a visual flowchart!

Challenges I ran into

In a couple occasions, I tried to shy away from yes/no answers to provide a more conversational flow, but I found that the flowchart logic kind of depends on them. Nonetheless, I believe the yes/no flow will provide users with a simple to follow experience.

Accomplishments that I'm proud of

I spent the month racking my brain for life hack ideas, but none of my ideas really made life that much easier. Until this one, that is! As a data science student, I can sit at my desk and multitask, asking Alexa which algorithm I should use while doing some data exploration, or something like that. I am proud to bring this solution to others as well.

What I learned

I learned that good ideas lurk right around the corner! Also, if you get stuck while trying to think of ideas, you may be thinking too hard! Just live your day-to-day life, and ideas may naturally emerge.

What's next for Machine Learning Algorithm Chooser

I have included a few sound effects in the skill to improve the user experience, but I'm considering fleshing them out and improving them in the future.

Built With

  • storyline
Share this project:

Updates