Related Experiment Video
Updated: Jan 4, 2026
06:19
Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
Published on: August 16, 2024
785
Using adaptive neuro-fuzzy inference system and multiple linear regression to estimate orange taste
Marzieh Mokarram1, Hosein Amin2, Mohammad R Khosravi3
1Department of Range and Watershed Management, College of Agriculture and Natural Resources of Darab Shiraz University Shiraz Iran.
Food Science & Nutrition
|October 30, 2019
Summary
Predicting orange taste using fruit characteristics like vitamin C and color is possible. The adaptive neuro-fuzzy inference system (ANFIS) demonstrated higher accuracy than multiple linear regression (MLR) for taste prediction.
Area of Science:
- Agricultural Science
- Food Science
- Data Science
Background:
- Orange taste is a crucial quality attribute for consumers.
- Objective prediction of orange taste can aid in quality control and breeding programs.
Purpose of the Study:
- To predict the taste of oranges using physical and chemical properties.
- To compare the predictive performance of Multiple Linear Regression (MLR) and Adaptive Neuro-Fuzzy Inference System (ANFIS).
Main Methods:
- Collected data on vitamin C, acid content, weight, diameter, and RGB color values for 70 orange samples.
- Applied MLR and ANFIS models to predict orange taste.
- Evaluated model performance using Mean Square Error (MSE) and Coefficient of Determination (R²).
Main Results:
- MLR showed significant relationships between taste and vitamin C, red, and blue color values.
- ANFIS achieved high accuracy with low error for training and checking data.
- ANFIS demonstrated a higher success rate in taste determination compared to MLR.
Conclusions:
- Fruit characteristics, particularly vitamin C and color, are strong predictors of orange taste.
- ANFIS is a more effective model than MLR for predicting orange taste.
- This study provides valuable insights for objective orange taste assessment.
