This project was based purely on the main intention to learn a new tool and have some fun.
Are chefs really to blame for the food you think tastes bad or good? Now we let the data talk.
What it does
Determine if fashion choices make a difference to food ratings.
How I built it
Using Mage's online tool, Kaggle datasets and some ML knowledge to train a good model..
Challenges I ran into
The fashion data is not complex enough to determine specific clothing article impact. It is likely that training the model on the serious option over what Mage calls the "intuitive" option would have resulted in a more accurate model. I definitely should have removed the component rating fields from training.
Accomplishments that I'm proud of
Tried a new tool. Focused entirely on learning this Hack. Compounded learning during a weekend filled with too many other commitments.
Initially I joined a team, but I knew I wouldn't be able to commit fully to the time required for this hackathon. So I decided to be honest to them, and even after they were cool with it – it didn't feel right with me to simply ride on their efforts. So I withdrew from their team before the hackathon started. And I'm excited to see what they cooked up : )
What I learned
Mage online portal. Practice story telling and laughter suppression.
What's next for How Fashion plays into our Food Bias
Use deployed API and integrate into webapp.