[Paper Review] OBESEYE: Interpretable Diet Recommender for Obesity Management using Machine Learning and Explainable AI

Mrinmoy Roy, Srabonti Das+1 more|arXiv (Cornell University)|Aug 5, 2023

Nutritional Studies and Diet8 citations

TL;DR

OBESEYE proposes an interpretable diet recommender that uses machine learning to predict individual nutrient requirements (carbohydrate, protein, fat, and fluid) and supports obesity management while accounting for comorbidities.

ABSTRACT

Obesity, the leading cause of many non-communicable diseases, occurs mainly for eating more than our body requirements and lack of proper activity. So, being healthy requires heathy diet plans, especially for patients with comorbidities. But it is difficult to figure out the exact quantity of each nutrient because nutrients requirement varies based on physical and disease conditions. In our study we proposed a novel machine learning based system to predict the amount of nutrients one individual requires for being healthy. We applied different machine learning algorithms: linear regression, support vector machine (SVM), decision tree, random forest, XGBoost, LightGBM on fluid and 3 other major micronutrients: carbohydrate, protein, fat consumption prediction. We achieved high accuracy with low root mean square error (RMSE) by using linear regression in fluid prediction, random forest in carbohydrate prediction and LightGBM in protein and fat prediction. We believe our diet recommender system, OBESEYE, is the only of its kind which recommends diet with the consideration of comorbidities and physical conditions and promote encouragement to get rid of obesity.

Motivation & Objective

• Motivate the need for personalized diet plans in obesity management, especially for patients with comorbidities.
• Develop a machine learning system to predict individual nutrient requirements (fluid, carbohydrate, protein, fat).
• Assess multiple ML models to identify which perform best for each nutrient prediction.
• Incorporate interpretability to support clinical decision-making and user trust.
• Promote diet recommendations that consider physical conditions and comorbidities.

Proposed method

• Apply multiple ML algorithms (linear regression, SVM, decision tree, random forest, XGBoost, LightGBM) to predict nutrient intake for fluid, carbohydrate, protein, and fat.
• Evaluate models using RMSE as the accuracy metric (low RMSE indicates better predictions).
• Identify the best-performing model per nutrient (e.g., linear regression for fluid, random forest for carbohydrate, LightGBM for protein and fat).
• Emphasize interpretability to align predictions with clinical relevance and explainability in dietary recommendations.

Experimental results

Research questions

• RQ1 Can machine learning accurately predict individual nutrient requirements for obesity management?
• RQ2 Which ML models are most effective for predicting fluid, carbohydrate, protein, and fat intake?
• RQ3 How does incorporating comorbidity and physical condition information influence diet recommendations?
• RQ4 Can the system generate interpretable diet recommendations that support clinical decision-making?

Key findings

• Linear regression achieved best performance for fluid prediction among the evaluated models.
• Random forest emerged as the best predictor for carbohydrate intake.
• LightGBM provided the strongest predictions for protein and fat intake.
• The study reports high predictive accuracy with low RMSE across the selected nutrients.
• OBESEYE is presented as a unique diet recommender that incorporates comorbidities and physical conditions in the recommendations.