2024, issue 2, p. 67-73
Received 09.04.2024; Revised 12.05.2024; Accepted 28.05.2024
Published 09.06.2024; First Online 14.06.2024
https://doi.org/10.34229/2707-451X.24.2.7
Previous | FULL TEXT (in Ukrainian) | Next
Using Support Vector Machine for Determining the Need for Artificial Watering Based on the Chlorophyll Fluorescence Induction
Volodymyr Hrusha 
V.M. Glushkov Institute of Cybernetics of the NAS of Ukraine, Kyiv
Correspondence: This email address is being protected from spambots. You need JavaScript enabled to view it.
Introduction. Chlorophyll fluorescence induction (CFI) is a method for receiving additional information about state of plants without its injuring. CFI occurs when the plant is illuminated in the blue spectrum of light. The so-called Kautsky curve in the red spectrum of light is measuring. Currently, there are scientific papers about determining water deficit using CFI and neural networks. Support Vector Machin (SVM) – the model of machine learning capable of performing linear or non-linear classification, regression and even finding outliers in data. SVM can be an alternative to neural networks for analysis of CFI measurements.
The purpose of the paper is to assess the possibility of determining the need for artificial watering of soybean plants based on the SVM and compare the results with the results obtained by the author in earlier experiments of analyzing the CFI of zinnia plants.
Results. SVM research was conducted using different SVM kernels, different methods of normalizing CFI measurements, different methods of the dimensionality reduction of CFI data. The SVM implementation from the Scikit-learn Python library, the SVR (Support Vector Regression) class was used. The best kernel of SVM, the best normalization method, and the best method of forming the input vector for the support vector machine were experimentally revealed for determining the need of artificial watering.
Conclusions. The research of Support Vector Machine for the purpose of determining the need for artificial watering based on CFI curves were conducted. The research showed that the best result is obtained (1) using a polynomial kernel of the fourth degree, (2) using 10 points of CFI curve taken unevenly on an exponentiation scale (power of 1/8), (3) using minimax normalization of CFI measurements. SVM showed a worse result in the analysis of CFI curves of soybean plants than in the analysis of CFI curves of the zinnia plants. It can be explained by the fact that the soybean is a drought resistant plant and therefore CFI demonstrated worse a moisture deficit.
Keywords: Support Vector Machine, chlorophyll fluorescence induction, dimensionality reduction of data.
Cite as: Hrusha V. Using Support Vector Machine for Determining the Need for Artificial Watering Based on the Chlorophyll Fluorescence Induction. Cybernetics and Computer Technologies. 2024. 2. P. 67–73. (in Ukrainian) https://doi.org/10.34229/2707-451X.24.2.7
References
1. Kalaji H.M, Schansker G., Ladle R.J. et al Frequently asked questions about in vivo chlorophyll fluorescence: practical issues. Photosynthesis Research. 2014. Vol. 122. Iss. 2. P. 121–158. https://doi.org/10.1007/s11120-014-0024-6
2. Kalaji H.M. et al. Frequently asked question about chlorophyll fluorescence, the sequel. Photosynthesis Research. 2017. Vol. 132. Iss. 1. P. 13 – 66. https://doi.org/10.1007/s11120-016-0318-y
3. Goltsev V., Zaharieva I., Chernev P., Kouzmanova M., Kalaji H.M., Yordanov I., Krasteva V., Alexandrov V., Stefanov D., Allkhverdiev S.I., Strasser R.J. Drought-induced modification of photosynthetic electron transport in intact leaves: Analysis and use of neural network as a tool for a rapid non-invasive estimation. Biochimica et Biophysica Acta. 2012. Vol. 1817. Iss. 8. P. 1490–1498. https://doi.org/10.1016/j.bbabio.2012.04.018
4. Zhou Ch., Mao J.D., Zhao H., Zhang B. Combining principle component analysis and a back-propagation network for classifying plant water stress based on images of chlorophyll fluorescence. Frontier research and Innovation in Optoelectronics technology and Industry. London: Taylor&Francis Group, 2019. P. 241–248.
5. Rybka K., Janaszek-Mankowska M., Siedlarz P., Mankowski D. Machine learning in determination of water saturation deficit in wheat leaves on basis of Chl a fluorescence parameters. Photosynthetica. 2019. Vol. 57. Iss. 1. P.226–230. http://dx.doi.org/10.32615/ps.2019.017
6. Hrusha V. Intelligent processing of data from chlorophyll fluorometric sensors. Cybernetics and computer technologies. 2022. Iss. 1. P. 42–48. (In Ukrainian) https://doi.org/10.34229/2707-451X.22.1.5
7. Srivastava D.K., Bhambhu L. Data classification using support vector machine. Journal of Theoretical and Applied Information Technology. 2010. Vol. 12. No. 1. https://www.jatit.org/volumes/Vol12No1/1Vol12No1.pdf
8. Scikit-learn: sklearn.svm. SVC. https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html (accessed: 25.03.2024).
9. Scikit-learn: sklearn.svm. SVR. https://scikit-learn.org/stable/modules /generated/sklearn.svm.SVR.html#sklearn. svm.SVR (accessed: 25.03.2024).
10. NumPy: numpy.poly. https://numpy.org/doc/stable/reference/generated/numpy.poly.html (accessed: 25.03.2024)
ISSN 2707-451X (Online)
ISSN 2707-4501 (Print)
Previous | FULL TEXT (in Ukrainian) | Next