MatchMaker

• 

  GUI

Inspiration

I found it hard to find a team member during this datathon. The process of finding teammates feels like searching for a needle in the haystack. MatchMaker - the team matching assistant is a perfect solution.

What it does

It compares the applicant's characteristics with other applicants in the hackathon and shortlists a group of best teammates.

How I built it

By preprocessing the data and visualizing patterns, I identified relevant features and then clustered them using K-means clustering algorithm. I evaluated it by checking the 2D and 3D PCA plots. Finally, I made it an application by building a GUI.

Challenges I ran into

Finding best visualizing ideas

Accomplishments that I'm proud of

Building an end-to-end product.

What I learned

I never worked with clustering algorithms before, this was a great chance to bring my theory knowledge of PCA and clustering to a good use

What's next for MatchMaker

Adding external data from other hackathons, social networking sites will make this app much more reliable and accurate.

Built With

• colab
• css
• flask
• html
• python
• scikit-learn