Updated: Sep 2, 2021
During the pandemic, many people have seemed to not discern a major problem apart from the disease. Many are disregarding the fact to eat good food at natural times. These people (admittedly including me) are not paying attention to the food they eat and at what time, deteriorating their health.
My solution to this? I have made an application that can recommend food to any user but with a twist: In my app, the most unique function is that the app personalizes the dataset for each user!
How you may ask? The application enters in a training phase by questioning the user’s daily habits and tastes. This information is added to the database, so it can be used for training. Also, this data is made a higher priority against the other generic data. Another interesting feature: after a cap limit of choice (default to 1000 rows), every time a new feature or input is added, the oldest row is thrown away. This means, that as the user changes his/her taste, the AI changes with them. As stated, the cap can also be changed based on the user’s preference.
Along with the main app is also some logs the user can refer to. The main one is a program called plot.py. This program shows the user the distribution of different categories in the dataset. This is useful to find out what trends the user has, and also to help the AI become more accurate by giving more examples in the necessary spot.
The dataset is always changing! One thing I hope to improve upon is to add a full-day feature (what meals can one afford to eat in a day), including money cap, calorie cap, and mood. This can also take into account events, such as birthday parties, or formal dinners with co-workers or employees.
All in all: this program can help anyone who wants to benefit from the stay at home and choose to eat healthier. Thanks!