Origin traceability of rice based on an electronic nose coupled with a feature reduction strategy

Abstract

Effective information processing technology is one of the keys to improving detection accuracy. In this study, a feature reduction strategy is proposed for reducing the dimension of electronic nose (e-nose) sensor features, in combination with multiclassifiers to identify the origin of rice. Firstly, the time domain and time-frequency domain features were extracted from the detection data. Secondly, the kernel principal component analysis and kernel entropy component analysis (KECA) were introduced to reduce the dimension of the fusion features to obtain the kernel principal components (KPCs) and kernel entropy components (KECs). Finally, global discriminant analysis (GDA) was proposed in order to reduce the dimension of the KPCs and KECs to obtain the final features, respectively. The results indicated that the KECA-GDA achieved the dimensionality reduction of fusion features, effectively, the good classification accuracy of 97% and 93.29%, F 1-scores of 0.9697 and 0.9410, and Kappa coefficients of 0.9648 and 0.9210 were obtained by means of the random forest (RF) method in uncooked and cooked rice, respectively. This study shows that KECA-GDA-RF can be used as an effective tool in tracing the origin of rice. Moreover, it can provide a useful processing technique to improve the measurement accuracy of an e-nose.

A global extended extreme learning machine combined with electronic nose for identifying tea gas information

“…It uses Renyi entropy to represent the feature structure of input space, ensure the minimum information loss in dimension reduction, and retain the maximum information entropy of features. 12,30 KECA is more conducive to finding the feature direction after dimensionality reduction.…”

“…Jia et al 11 proposed a multi-core support vector machine algorithm to identify the pollutant gas based on e-nose technology. Shi et al 12 tracked the origin of rice based on the multi-recognition technology, and the difference of gas features was visualized. Wang et al 13 proposed a transfer learning method combined with e-nose to identify the xihu longjing tea.…”

“…Compared to SVM and RF, ELM obtained the best classification performance. 15 Shi et al 12 obtained the gas information of rice from different origins based on the e-nose technology and traced the origin of the gas information of the rice based on the ELM algorithm. Lahoura et al obtained online data of breast cancer based on cloud technology and combined it with ELM for online identification.…”

Recent Progress in Smart Electronic Nose Technologies Enabled with Machine Learning Methods

“…Various hyperparameters were tested including the width and the type of window, of which a training window of size 480 s produced the best classification accuracy. Frequency-domain features can be extracted after transforming the original time-domain signal to a frequency domain through Fourier transform or wavelet transform [47][48][49][50][51]. During the experiment by Dai et al [6], who used MOX sensors to classify different teas, original signals were transformed into a vector by wavelet packet decomposition with Daubechies wavelet as the wavelet base.…”

Feature Reduction Method Coupled with Electronic Nose for Quality Control of Tea

“…The signal processing performs feature extraction and processing to remove redundant information, and the pattern recognition makes the classification decision. Owing to its use of sensor detection technology, the e-nose has the advantages of high stability, rapid processing, and simple operation, and it has been widely used in food engineering, (3,4) electrical engineering, (5) and medical engineering. (6,7) After obtaining the detection data, a feature reduction method will affect the detection performance of the e-nose.…”