udayan123 / diet-recommendation-system-

Introduction

A wide variety of ingredients, cultures and personal tastes makes decisions about what to eat a great problem. Many diseases that were previously thought of as hereditary are now seen to be connected to biological dysfunction related to nutrition. Being healthy and eating better is something the vast majority of the population wants and doing so usually requires great effort. The working prototype accomplishes a Personalized Diet Recommendation System with the integration of Machine Learning Algorithms to recommend the right food at right time and with the right nutrition, calories, fat, etc.

Objective

To establish a working prototype of a Personalized Diet Recommendation System.

How we built it

Design Architecture

Model Design

Design Architecture

Challenges we ran into

The issue that is being faced by the current working model is it does not facilitate a user with the food classification based on the food timings on a daily basis other than general food timings. The existing system struggles to provide a weight gain/loss scheme to a user based on his long-term food habits.

Accomplishments that we're proud of

I learned a lot of different Machine Learning algorithms along with how to build a Recommendation System.

Results

A working prototype of a Diet Recommendation System is established. The module works on the basis of K-Means Clustering and Random Forest Classification Algorithms. Tkinter-based GUI is implemented.

What we learned

• Tkinter
• Recommendation System
• Random Forest
• K-Mean Clustering

What's next for Diet Recommendation System in Healthcare using ML Approaches

The module can be implemented as a cloud-based application. Packaged as a single entity, ready for production environment deployment.