Data Mining Models for Selection of the Best Spectral Reflectance Indices in Estimation of Crop Yields and Classification of Maize Hybrid Types Using SpectroRadiometer Data

EROL H., BARUTÇULAR C., El Sabagh A., Erol R.

European Conference on Electrical Engineering and Computer Science (EECS), Bern, Switzerland, 17 - 19 November 2017, pp.180-187 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/eecs.2017.42
  • City: Bern
  • Country: Switzerland
  • Page Numbers: pp.180-187
  • Çukurova University Affiliated: Yes


This study purposes data mining models to estimate the amounts of crop yields using the relationships between the numeric valued crop yield attributes and the numeric valued spectral reflectance indices attributes calculated using different range of canopy reflectance. Data mining models uses knowledge and data technology to find the best spectral reflectance indices subset selection in estimation of crop yields for spectroradiometer reflectance measurements in 220 nm to 1100 nm range. Crop traits are estimated by use of linear regression models as data mining models in terms of computed values of spectral reflectance indices. Data mining classification method with high performance algorithm is used to classify different types of maize hybrids using the numeric valued crop yield attributes with respect to the nominal valued attributes corresponding to different conditions in this study.