Our repo: https://github.com/Datathon2023TeamDNF/MangoChallenge2

Inspiration

In order to tackle the challenge, we took advantage of our experience in competitive programming competitions, developing a novel approach to the task based of graph theory concepts.

What it does

Our system takes a piece of clothing from the user and builds a matching outfit,.

How we built it

We took full advantage of the dataset the challenge provides, harnessing its graph structure in order to build new relationships between clothing items.

Challenges we ran into

The development and integration of the output generation pipeline, encompassing graph generation, inference based on user inputs and the construction of the user interface, proved a challenging issue which stretched our skills.

Accomplishments that we're proud of

We feel accomplished in the development of a novel method to treat this problem, using a graph instead of other more common deep learning approaches.

What we learned

During the challenge, we used a wide variety of coding skills, which we feel has made us better coders.

What's next for Mango Challenge - Team DNF

As a team, we look forward to new challenges as interesting as Mango's proposal, in order to help us further hone our capabilities.

Built With

• c++
• pandas
• python
• streamlit