TasteBuds

Your pheno-lized Yelp listings!

Inspiration

We wanted to see if genome data can help us find specific Yelp listings that we may like intrinsically.

What it does

With our heuristic, we take the genome data input it into our predictive model, which invokes a Yelp search with high ranked Yelp categories. We integrated Google Maps into our app to give the user a look at where those Yelp listings around them.

How we built it

We built upon the Flask example app and implemented the API integration.

Challenges we ran into

Feature confusion and low data, as well as high bias in our weights.

Accomplishments that we're proud of

We're proud that the Yelp listings we see at least pass our sanity checks.

What we learned

Look at the data and check the slack for more details! Always pivot sooner than later.

What's next for sdHacks17

Improve on our full-stack and data science/engineering skills with more projects like this!

Built With

Submitted to

SD Hacks 2017

Created by

I developed the API integration, data cleaning, feature selection and modeling heuristic. This project was from my ideation.

Tommy Truong
I designed and implemented the model for suggesting categories based off phenotype scores.

George Alfaro
Team biology and genetics expert and bio-data analyst.

Dylan Mills
I worked on the backend, created the genomelink api pull and json parsing methods to generate a list of phenotype categories and scores. The phenotype to yelp search category mapping method was my brainchild.

Joshua Hernandez