The Use of Gradient Boost Regression Model to Modeling of Gas Sensors in Diagnosis of Sun-dried, Sulphurous and Acidic solution dried Raisins

Document Type : Research Paper

Authors

1 PhD student, Department of Agricultural Machinery Engineering, Faculty of Agriculture, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran

2 Professor, Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University College of Agriculture and Natural Resources, University of Tehran, Karaj, Iran

3 Professor, Department of Agricultural Machinery Engineering, Faculty of Agriculture, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran

Abstract

Machine learning modeling can help overcome some of the limitations of gas sensors, such as high operational conditions, drift errors, limited selectivity, the need for a large amount of labeled data, and cost and fabrication challenges. In this research, an electronic nose system was developed for the detection of sulfur dioxide and acetic acid. three treatments, including sunny, acetic, and sulfuric, were prepared in three repetitions, and each was exposed to olfactory sensors for 60 minutes to record the sensor responses to each treatment. Then, the data obtained from the sensor responses were examined by machine learning models to determine the modeling accuracy of each method. The results showed that the utilized Gradient Boost Regression model with a determination coefficient of 0.9972, root mean square error of 0.0209, mean absolute error of 0.0026, and relative root mean square error of 0.0209 was able to model the gas sensor responses well for the introduced treatments. Furthermore, by analyzing the results, the type and degree of correlation between the sensor responses to each other and over time were determined to evaluate their behavior prediction. Then, based on the conducted modeling, it was revealed that MQ9, MQ3, MQ5, and TGS2620 sensors, with determination coefficients of 0.8668, 0.8786, 0.9458, and 0.9074, and root mean square errors of 0.0163, 0.0168, 0.0083, and 0.0227, respectively, provided more accurate and predictable responses compared to MQ135, TGS822, TGS810, and MQ4 sensors.

Keywords

Main Subjects


The Use of Gradient Boost Regression Model to Modeling of Gas Sensors in Diagnosis of Sun-dried, Sulphurous and Acidic solution dried Raisins

EXTENDED ABSTRACT

Introduction

Remaining elements such as sulfur dioxide and its compounds, which are widely used in dried fruits as preservatives due to their availability and affordability, are one of the factors that importing countries consider when purchasing raisins from Iran. Therefore, exported raisins should be examined for the presence of these elements in the final product. Machine learning modeling can help overcome some of the limitations of gas sensors, such as high operating conditions, drift errors, limited selectivity, the need for a large amount of labeled data, and cost and manufacturing challenges. The gradient boost regression model is a machine learning model used to solve regression problems.

Materials and Methods

In this study, three treatments, including sun-dried, Acidic solution dried Raisins and sulfur-treated each with three replicates, were prepared and exposed to the gas sensors for 60 minutes to record the sensor responses to each treatment. The obtained data were then analyzed using machine learning models to determine the accuracy of each modeling method and make them comparable. The model evaluation parameters were examined, and the interpretation of each was discussed in detail. Finally, the analysis of variance of the gradient boost regression model was performed for each quality prediction component separately for treatments, sensors, and combinations of sensors with treatments, and various points were extracted from the interpretation of each in the discussion and results section.

Results and Discussion

Based on the charts and results, the gradient boost regression model has been able to provide more accurate and better predictions in all sensors. Therefore, the modeling by this model with the quality determining components of the model was analyzed and the modeling results were examined. Overall, considering the high values of quality prediction metrics, it can be concluded that the designed gradient boost regression model is well compatible with the dataset and can effectively predict the target variable. The significant difference test results also showed significant differences between the mean treatments. Treatments 1 to 3, corresponding to acidic solution dried raisins, sulfuric-treated raisins, and sun-dried raisins, were found to have significant differences. According to the results, the coefficient of determination in the Acidic solution treatment had a significant difference compared to the Sun-dried and sulfuric treatments, and performed better than both. Additionally, the Sun-dried treatment had a significant difference compared to the solar treatment, and the results of the solar treatment were better, indicating that the Acidic solution treatment had the highest modeling capability and predictability of sensor responses, which can be justified by the more noticeable odor created by acid. Furthermore, the Sun-dried treatment showed the lowest modeling capability compared to other treatments, which can be justified by the lack of clear processing performed on it compared to other treatments. However, no significant differences were observed in the root mean squared error, mean absolute error, and root mean squared error of relative error between the mean treatments. Comparing the means of responses of each sensor also showed that for the comparisons of sensors 1 and 2, 1 and 3, 2 and 3, 2 and 5, 4 and 5, and 6 and 7, the reject value was False, indicating that the null hypothesis was accepted, meaning that there is no significant difference between these sensors. In the other comparisons, the reject value was True, indicating that the null hypothesis was rejected and there is a significant difference between them.

Conclusion

The results showed that the gradient boost regression model with the coefficient of explanation of 0.9972 and the root mean square error of 0.0209 for the test data was able to model the response of the gas sensors compared to treatments. Also, by examining and analyzing the obtained results, the type and degree of correlation between the response of the sensors in relation to each other and in relation to time was determined to be evaluated in predicting their behavior. Then, with the modeling done, it was determined that the MQ9, MQ3, MQ5, TGS2620 sensors have coefficients of explanation of 0.8668, 0.8786, 0.9458, and 0.9074, respectively, and the root mean square error of 0.0163, 0.0168, and 0.0083. 0 and 0.0227 showed more accurate and predictable responses than MQ135, TGS822, TGS810 and MQ4 sensors.

Baha, H., & Dibi, Z. (2010). ANN modeling of a gas sensor. Journal of Electrical Engineering & Technology, 5(3), 493–496. https://doi.org/10.5370/jeet.2010.5.3.493.
Borowik, P., Adamowicz, L., Tarakowski, R., Siwek, K., & Grzywacz, T. (2020a). Odor detection using an E-Nose with a reduced sensor array. Sensors, 20(12), 3542. https://doi.org/10.3390/s20123542
European Food Safety Authority (EFSA). (2016). Panel on food additives and nutrient sources added to food (ANS). Scientific opinion on the re-evaluation of sulfur dioxide (E 220), sodium sulfite (E 221), sodium bisulfite (E 222), sodium metabisulfite (E 223), potassium metabisulfite (E 224), calcium sulfite (E 226), calcium bisulfite (E 227), and potassium bisulfite (E 228) as food additives. EFSA Journal. 14: 4438. [DOI: 10.2903/j.efsa.2016.4438]
Ghasemi‐Varnamkhasti, M., Mohammad‐Razdari, A., Yoosefian, S. H., Izadi, Z., & Rabiei, G. (2019). Selection of an optimized metal oxide semiconductor sensor (MOS) array for freshness characterization of strawberry in polymer packages using response surface method (RSM). Postharvest Biology and Technology, 151, 53–60. https://doi.org/10.1016/j.postharvbio.2019.01.016
Guido L.F. (2016). Sulfites in beer: reviewing regulation, analysis and role. Scientia Agricola. 73: 189-197. [DOI: 10.1590/0103- 9016-2015-0290].
https://doi.org/10.1016/j.jspr.2021.101805
Institute of Standards and Industrial Research of Iran (ISIRI). (2014). Gheisi (Whole dried apricot) specifications and test methods. National Standard No. 13. 4th revision. URL: http://standard.isiri.gov.ir/StandardView.aspx?Id=41629. Accessed 7 June 2014.
Institute of Standards and Industrial Research of Iran (ISIRI). (2015). Dried apricots - specification and test methods. National Standard No. 11. 5th revision. URL: gov.ir/StandardView.aspx?Id=40024. Accessed 22 November 2015.
Institute of Standards and Industrial Research of Iran (ISIRI). (2017). Dried fruits- determination of sulfur dioxide. National Standard No. 569. URL: http://standard.isiri.gov.ir/ StandardView.aspx?Id=47184. Accessed 25 December 2017.
Institute of Standards and Industrial Research of Iran (ISIRI). (2018). Specification and methods of test for fruit snack (fruit paste). National Standard No. 3308. 2nd revision. URL: http://standard.isiri.gov.ir/StandardView.aspx?Id=50065. Accessed 25 August 2018.
Jamalizadeh, F., Ghasemi-Varnamkhasti, M., Ghasemi Nafchi, M., Tohidi, M., & Dowlati, M. (2020). Implementation of an olfactory machine system for the classification of different types of black pepper based on geographical origin and detection of cheating in Indian black pepper. Iranian Food Science and Technology Research Journal, 16(4), 479-491.
Keramat-Jahromi, M., Mohtasebi, S. S., Mousazadeh, H., Ghasemi-Varnamkhasti, M., rafiee, S., & Savand-Roumi, E. (2019). Evaluation of a Machine Olfaction to Classify the Quality of Dried Date Fruit by Electrohydrodynamic, Hot Air, and the Hybrid Drying Techniques. Iranian Journal of Biosystems Engineering, 50(1), 241-251. doi: 10.22059/ijbse.2018.248873.665023.(In Persian).
Kheireddine, Lamamra., Djamil, Rechem. (2016). Artificial neural network modelling of a gas sensor for liquefied petroleum gas detection.   doi: 10.1109/ICMIC.2016.7804292
Khodamoradi, F., Mirzaee‐Ghaleh, E., Dalvand, M. J., & Sharifi, R. (2021). Classification of basil plant based on the level of consumed nitrogen fertilizer using an olfactory machine. Food Analytical Methods, 14(12), 2617–2629. https://doi.org/10.1007/s12161-021-02089-y
Kiani, S., Minaei, S., & Ghasemi-Varnamkhasti, M. (2017). Integration of computer vision and electronic nose as non-destructive systems for saffron adulteration detection. Computers and Electronics in Agriculture, 141, 46–53. https://doi.org/10.1016/j.compag.2017.06.018
Leon-Medina, J. X., Pineda-Muñoz, W. A., & Burgos, D. a. T. (2020). Joint distribution adaptation for drift correction in electronic NOSE type sensor arrays. IEEE Access, 8, 134413–134421. https://doi.org/10.1109/access.2020.3010711 Ye, Z., Liu, Y., & Li, Q. (2021). Recent Progress in Smart Electronic Nose Technologies Enabled with Machine Learning Methods. Sensors, 21(22), 7620. https://doi.org/10.3390/s21227620
Liu, Y., Wang, Q., Xu, Q., Feng, J., Yu, H., & Yin, Y. (2018). Non-destructive detection of Flos Lonicerae treated by sulfur fumigation based on hyperspectral imaging. Journal of Food Measurement and Characterization, 12(4), 2809-2818.
Liu, Y., Wang, Q., Xu, Q., Feng, J., Yu, H., & Yin, Y. (2018b). Non-destructive detection of Flos Lonicerae treated by sulfur fumigation based on hyperspectral imaging. Journal of Food Measurement and Characterization, 12(4), 2809–2818. https://doi.org/10.1007/s11694-018-9896-z
Maho, P., Herrier, C., Livache, T., Comon, P., & Barthelmé, S. (2022). A calibrant-free drift compensation method for gas sensor arrays. Chemometrics and Intelligent Laboratory Systems, 225, 104549. https://doi.org/10.1016/j.chemolab.2022.104549
Makarichian, A., Chayjan, R. A., Ahmadi, E., & Mohtasebi, S. S. (2021). Assessment the influence of different drying methods and pre-storage periods on garlic (Allium Sativum L.) aroma using electronic nose. Food and Bioproducts Processing, 127, 198–211. https://doi.org/10.1016/j.fbp.2021.02.016
Men, H., Shi, Y., Jiao, Y., Gong, F., & Liu, J. (2018). Electronic nose sensors data feature mining: a synergetic strategy for the classification of beer. Analytical Methods, 10(17), 2016–2025. https://doi.org/10.1039/c8ay00280k
Meng Z., Qin G., Zhang B. (2005). DNA damage in mice treated with sulfur dioxide by inhalation. Environmental and Molecular Mutagenesis. 46: 150-155. [DOI: 10.1002/em.20142].
Mischek D., Krapfenbauer-Cermak C. (2012). Exposure assessment of food preservatives (sulphites, benzoic and sorbic acid) in Austria. Food Additives and Contaminants: Part A. 29: 371-382. [DOI: 10.1080/19440049.2011.643415]
Mohammad‐Razdari, A., Ghasemi‐Varnamkhasti, M., Yoosefian, S. H., Izadi, Z., & Siadat, M. (2019). Potential application of electronic nose coupled with chemometric tools for authentication assessment in tomato paste. Journal of Food Process Engineering, 42(5). https://doi.org/10.1111/jfpe.13119
Ordukaya, E., & Karlık, B. (2017). Quality control of olive oils using machine learning and electronic nose. Journal of Food Quality, 2017, 1–7. https://doi.org/10.1155/2017/9272404
Ozgoli, H., Mohtasebi, S. S., Hosseinpour, S., & Hosseinpour-Zarnaq, M. (2023). Investigating meat and oil quality in chicken nuggets using electronic nose and image processing techniques. Iranian Journal of Biosystems Engineering, 54(2), 1-14. doi: 10.22059/ijbse.2023.363744.665517.(In Persian).
Rahimzadeh, H., Sadeghi, M., Mireei, S. A., & Ghasemi‐Varnamkhasti, M. (2022). Unsupervised modelling of rice aroma change during ageing based on electronic nose coupled with bio-inspired algorithms. Biosystems Engineering, 216, 132–146. https://doi.org/10.1016/j.biosystemseng.2022.02.010
Soubra L., Sarkis D., Hilan C., Verger P. (2007). Dietary exposure of children and teenagers to benzoates, sulphites, butylhydroxyanisol (BHA) and butylhydroxytoluen (BHT) in Beirut (Lebanon). Regulatory Toxicology and Pharmacology. 47: 68-77. [DOI: 10.1016/j.yrtph.2006.07.005].
Souhil, Kouda., T., Bendib., Samir, Barra., Abdelghani, Dendouga. (2018). ANN modeling of an industrial gas sensor behavior.   doi: 10.1109/CCEE.2018.8634510
Tahri, K., Bari, N. E., & Bouchikhi, B. (2016). Geographical classification and adulteration detection of cumin by using electronic sensing coupled to. . . ResearchGate. https://www.researchgate.net/publication/346355867_Geographical_classification_and_adulteration_detection_of_cumin_by_using_electronic_sensing_coupled_to_multivariate_analysis
Tohidi, M., Ghasemi-Varnamkhasti, M., Ghafarinia, V., Mohtasebi, S. S., & Bonyadian, M. (2018). Identification of trace amounts of detergent powder in raw milk using a customized low-cost artificial olfactory system: A novel method. Measurement, 124, 120–129. https://doi.org/10.1016/j.measurement.2018.04.006
Vally H., Misso N.L.A., Madan V. (2009). Clinical effects of sulphite additives. Clinical and Experimental Allergy. 39: 1643-1651. [DOI: 10.1111/j.1365-2222.2009.03362.x].
Yang, Y., Shahbeik, H., Shafizadeh, A., Masoudnia, N., Rafiee, S., Zhang, Y., Pan, J., Tabatabaei, M., & Aghbashlo, M. (2022). Biomass microwave pyrolysis characterization by machine learning for sustainable rural biorefineries. Renewable Energy, 201, 70–86. https://doi.org/10.1016/j.renene.2022.11.028.
Zhou, M., Khir, R., Pan, Z., Campbell, J. F., Mutters, R., & Hu, Z. (2021). Feasibility of detection of infested rice using an electronic nose. Journal of Stored Products Research, 92, 101805.