Fusion of machine learning and explainable AI for enhanced rice classification: a case study on Cammeo and Osmancik species

IF 3 3区 农林科学 Q2 FOOD SCIENCE & TECHNOLOGY
Ahmet Çifci, İsmail Kırbaş
{"title":"Fusion of machine learning and explainable AI for enhanced rice classification: a case study on Cammeo and Osmancik species","authors":"Ahmet Çifci,&nbsp;İsmail Kırbaş","doi":"10.1007/s00217-024-04614-9","DOIUrl":null,"url":null,"abstract":"<div><p>The accurate identification and classification of rice species is critical for increasing crop productivity, quality, and diversity. Traditional rice classification methods involving manual inspection can be time-consuming, costly, and error-prone. This study addresses to address this challenge by exploring the potential of machine learning (ML) models for automated and accurate rice classification. The key objectives of this paper are threefold. First, the study evaluates the discriminative power of various morphological features extracted from rice grain images using feature selection methods. Second, it compares the performance of several ML models, including Artificial Neural Network (ANN), Categorical Boosting (CatBoost), Gradient Boosting (GBoost), k-Nearest Neighbours (k-NN), Logistic Regression (LR), Naïve Bayes (NB), Random Forest (RF), Stochastic Gradient Descent (SGD), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost), in classifying two rice species (Cammeo and Osmancik). Third, the study implements explainable artificial intelligence (XAI) techniques, namely SHapley Additive exPlanation (SHAP) and Individual Conditional Expectation (ICE) plots, to provide transparency and interpretability into the inner workings and decision-making processes of the ML models. The findings indicate that the LR model achieved the highest classification accuracy, with a rate of 93.1%. Feature analysis identified Major Axis Length, Perimeter, Convex Area, and Area as the most influential features in distinguishing between rice species. This study highlights the successful application of advanced ML techniques in automating industrial rice classification, facilitating automated packaging and quality control processes without the need for human intervention. By improving the efficiency of rice classification and reducing reliance on manual labour, this approach offers significant benefits to both the agricultural industry and food production sectors.</p></div>","PeriodicalId":549,"journal":{"name":"European Food Research and Technology","volume":"251 1","pages":"69 - 86"},"PeriodicalIF":3.0000,"publicationDate":"2024-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s00217-024-04614-9.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"European Food Research and Technology","FirstCategoryId":"97","ListUrlMain":"https://link.springer.com/article/10.1007/s00217-024-04614-9","RegionNum":3,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"FOOD SCIENCE & TECHNOLOGY","Score":null,"Total":0}
引用次数: 0

Abstract

The accurate identification and classification of rice species is critical for increasing crop productivity, quality, and diversity. Traditional rice classification methods involving manual inspection can be time-consuming, costly, and error-prone. This study addresses to address this challenge by exploring the potential of machine learning (ML) models for automated and accurate rice classification. The key objectives of this paper are threefold. First, the study evaluates the discriminative power of various morphological features extracted from rice grain images using feature selection methods. Second, it compares the performance of several ML models, including Artificial Neural Network (ANN), Categorical Boosting (CatBoost), Gradient Boosting (GBoost), k-Nearest Neighbours (k-NN), Logistic Regression (LR), Naïve Bayes (NB), Random Forest (RF), Stochastic Gradient Descent (SGD), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost), in classifying two rice species (Cammeo and Osmancik). Third, the study implements explainable artificial intelligence (XAI) techniques, namely SHapley Additive exPlanation (SHAP) and Individual Conditional Expectation (ICE) plots, to provide transparency and interpretability into the inner workings and decision-making processes of the ML models. The findings indicate that the LR model achieved the highest classification accuracy, with a rate of 93.1%. Feature analysis identified Major Axis Length, Perimeter, Convex Area, and Area as the most influential features in distinguishing between rice species. This study highlights the successful application of advanced ML techniques in automating industrial rice classification, facilitating automated packaging and quality control processes without the need for human intervention. By improving the efficiency of rice classification and reducing reliance on manual labour, this approach offers significant benefits to both the agricultural industry and food production sectors.

求助全文
约1分钟内获得全文 求助全文
来源期刊
European Food Research and Technology
European Food Research and Technology 工程技术-食品科技
CiteScore
6.60
自引率
3.00%
发文量
232
审稿时长
2.0 months
期刊介绍: The journal European Food Research and Technology publishes state-of-the-art research papers and review articles on fundamental and applied food research. The journal''s mission is the fast publication of high quality papers on front-line research, newest techniques and on developing trends in the following sections: -chemistry and biochemistry- technology and molecular biotechnology- nutritional chemistry and toxicology- analytical and sensory methodologies- food physics. Out of the scope of the journal are: - contributions which are not of international interest or do not have a substantial impact on food sciences, - submissions which comprise merely data collections, based on the use of routine analytical or bacteriological methods, - contributions reporting biological or functional effects without profound chemical and/or physical structure characterization of the compound(s) under research.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信