An ensemble incremental approach of Extreme Learning Machine (ELM) For paddy growth stages classification using MODIS remote sensing images

Remote sensing technology plays an important role in agriculture applications, especially for paddy growth stages classification, which is a critical process in predicting crop production. The analysis of multi bands data covering very large swath areas using iterative methods such as neural network or SVM will certainly cost much computation time. This paper addresses this problem by taking advantage from a non-iterative tuning capability of Extreme Learning Machine (ELM) for paddy growth stages classification using MODIS (Moderate Resolution Imaging Spectroradiometer) remote sensing images. The accuracy of classification is measured by Cohen's kappa. Seven classes are used in classification, with consist of six classes for paddy growth stages and one class for dominated cloud. The contribution of this study is a new ensemble incremental approaches based on random bootstrap resampling for basic ELM (B-ELM) and Error Minimized ELM (EM-ELM) are applied to build multi-class classifier using two types of hidden nodes function, i.e. additive and radial basis function (RBF) hidden nodes. The classification results were compared each other with these two types of hidden nodes. Our ensemble incremental approach successfully classify seven paddy growth stages and significantly improve the overall kappa coefficient to 10.2% higher with only in average 7 nodes addition overhead.

[1]  Chee Kheong Siew,et al.  Universal Approximation using Incremental Constructive Feedforward Networks with Random Hidden Nodes , 2006, IEEE Transactions on Neural Networks.

[2]  Lior Rokach,et al.  Data Mining And Knowledge Discovery Handbook , 2005 .

[3]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[4]  Mohamad Ivan Fanany,et al.  An Extreme Learning Machine Model for Growth Stages Classification of Rice Plants from Hyperspectral Images Subdistrict Indramayu , 2013 .

[5]  Thomas G. Dietterich.  Multiple Classifier Systems , 2000, Lecture Notes in Computer Science.

[6]  T. Basaruddin,et al.  A paddy growth stages classification using MODIS remote sensing images with balanced branches support vector machines , 2012, 2012 International Conference on Advanced Computer Science and Information Systems (ICACSIS).

[7]  OpitzDavid,et al.  Popular ensemble methods , 1999 .

[8]  Amaury Lendasse,et al.  Extreme learning machines for soybean classification in remote sensing hyperspectral images , 2014, Neurocomputing.

[9]  Lior Rokach,et al.  Ensemble Methods for Classifiers , 2005, The Data Mining and Knowledge Discovery Handbook.

[10]  Min Han,et al.  A Remote Sensing Image Classification Method Based on Extreme Learning Machine Ensemble , 2013, ISNN.

[11]  Sidik Mulyono,et al.  Genetic algorithm based new sequence of principal component regression (GA-NSPCR) for feature selection and yield prediction using hyperspectral remote sensing data , 2012, 2012 IEEE International Geoscience and Remote Sensing Symposium.

[12]  Robert K. L. Gay,et al.  Error Minimized Extreme Learning Machine With Growth of Hidden Nodes and Incremental Learning , 2009, IEEE Transactions on Neural Networks.

[13]  Hongming Zhou,et al.  Extreme Learning Machine for Regression and Multiclass Classification , 2012, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).