Supervised and Semi-Supervised Self-Organizing Maps for Regression and Classification Focusing on Hyperspectral Data

Machine learning approaches are valuable methods in hyperspectral remote sensing, especially for the classification of land cover or for the regression of physical parameters. While the recording of hyperspectral data has become affordable with innovative technologies, the acquisition of reference data (ground truth) has remained expensive and time-consuming. There is a need for methodological approaches that can handle datasets with significantly more hyperspectral input data than reference data. We introduce the Supervised Self-organizing Maps (SuSi) framework, which can perform unsupervised, supervised and semi-supervised classification as well as regression on high-dimensional data. The methodology of the SuSi framework is presented and compared to other frameworks. Its different parts are evaluated on two hyperspectral datasets. The results of the evaluations can be summarized in four major findings: (1) The supervised and semi-Supervised Self-organizing Maps (SOM) outperform random forest in the regression of soil moisture. (2) In the classification of land cover, the supervised and semi-supervised SOM reveal great potential. (3) The unsupervised SOM is a valuable tool to understand the data. (4) The SuSi framework is versatile, flexible, and easy to use. The SuSi framework is provided as an open-source Python package on GitHub.

[1]  Pedro H. M. Braga,et al.  A Semi-Supervised Self-Organizing Map for Clustering and Classification , 2018, 2018 International Joint Conference on Neural Networks (IJCNN).

[2]  Yoshifumi Nishio,et al.  Batch-Learning Self-Organizing Map with Weighted Connections avoiding false-neighbor effects , 2010, The 2010 International Joint Conference on Neural Networks (IJCNN).

[3]  O. Chapelle,et al.  Semi-Supervised Learning (Chapelle, O. et al., Eds.; 2006) [Book reviews] , 2009, IEEE Transactions on Neural Networks.

[4]  Kuolin Hsu,et al.  Self‐organizing linear output map (SOLO): An artificial neural network suitable for hydrologic modeling and analysis , 2002 .

[5]  Ah Chung Tsoi,et al.  A supervised training algorithm for self-organizing maps for structures , 2005, Pattern Recognit. Lett..

[6]  Liangpei Zhang,et al.  An unsupervised artificial immune classifier for multi/hyperspectral remote sensing imagery , 2006, IEEE Trans. Geosci. Remote. Sens..

[7]  Alexander Gepperth,et al.  Using self-organizing maps for regression: the importance of the output function , 2015, ESANN.

[8]  A. Ultsch,et al.  Self-Organizing Neural Networks for Visualisation and Classification , 1993 .

[9]  Alexander N. Gorban,et al.  SOM: Stochastic initialization versus principal components , 2016, Inf. Sci..

[10]  Stefan Pollmann,et al.  Neuroinformatics Original Research Article Pymvpa: a Unifying Approach to the Analysis of Neuroscientifi C Data , 2022 .

[11]  Sina Keller,et al.  Introducing a Framework of Self-Organizing Maps for Regression of Soil Moisture with Hyperspectral Data , 2018, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium.

[12]  Bernhard Schölkopf,et al.  Semi-Supervised Learning (Adaptive Computation and Machine Learning) , 2006 .

[13]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[14]  A. M. Kalteh,et al.  Review of the self-organizing map (SOM) approach in water resources: Analysis, modelling and application , 2008, Environ. Model. Softw..

[15]  Sheng-Hsun Hsu,et al.  A two-stage architecture for stock price forecasting by integrating self-organizing map and support vector regression , 2009, Expert Syst. Appl..

[16]  Edoardo Pasolli,et al.  Ensemble Multiple Kernel Active Learning For Classification of Multisource Remote Sensing Data , 2015, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[17]  Teuvo Kohonen,et al.  The self-organizing map , 1990, Neurocomputing.

[18]  Ali Selamat,et al.  An empirical study based on semi-supervised hybrid self-organizing map for software fault prediction , 2015, Knowl. Based Syst..

[19]  Felix M. Riese.  SUSI: SUpervised Self-organIzing maps in Python , 2019 .

[20]  Diego Vidaurre,et al.  SOMwise regression: a new clusterwise regression method , 2011, Neural Computing and Applications.

[21]  José Alberto Sá,et al.  Recurrent Self-Organizing Map for Severe Weather Patterns Recognition , 2012 .

[22]  Luis Gómez-Chova,et al.  Remote Sensing Image Processing , 2011, Remote Sensing Image Processing.

[23]  Patrice Aknin,et al.  Comparison of Supervised Self-Organizing Maps Using Euclidian or Mahalanobis Distance in Classification Context , 2001, IWANN.

[24]  Shai Ben-David,et al.  Understanding Machine Learning: From Theory to Algorithms , 2014 .

[25]  Panu Somervuo,et al.  How to make large self-organizing maps for nonvectorial data , 2002, Neural Networks.

[26]  Travis E. Oliphant,et al.  Python for Scientific Computing , 2007, Computing in Science & Engineering.

[27]  Stefan Hinz,et al.  Hyperspectral Data and Machine Learning for Estimating CDOM, Chlorophyll a, Diatoms, Green Algae and Turbidity , 2018, International journal of environmental research and public health.

[28]  Ron Wehrens,et al.  Flexible Self-Organizing Maps in kohonen 3.0 , 2018 .

[29]  Claudia Notarnicola,et al.  Review of Machine Learning Approaches for Biomass and Soil Moisture Retrievals from Remote Sensing Data , 2015, Remote. Sens..

[30]  Stefan Hinz,et al.  DEVELOPING A MACHINE LEARNING FRAMEWORK FOR ESTIMATING SOIL MOISTURE WITH VNIR HYPERSPECTRAL DATA , 2018, ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences.

[31]  R. Horowitz,et al.  Convergence properties of self-organizing neural networks , 1995, Proceedings of 1995 American Control Conference - ACC'95.

[32]  Giovanni Zurlini,et al.  Spectral Self-Organizing Map for hyperspectral image classification , 2003 .

[33]  Aluizio F. R. Araújo,et al.  Identification and control of dynamical systems using the self-organizing map , 2004, IEEE Transactions on Neural Networks.

[34]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[35]  Johannes R. Sveinsson,et al.  Random Forests for land cover classification , 2006, Pattern Recognit. Lett..

[36]  Sina Keller,et al.  Hyperspectral benchmark dataset on soil moisture , 2018 .

[37]  Esa Alhoniemi,et al.  Clustering of the self-organizing map , 2000, IEEE Trans. Neural Networks Learn. Syst..

[38]  Antonio Plaza,et al.  Hyperspectral Image Classification using a Self-Organizing Map , 2001 .

[39]  W. Natita,et al.  Appropriate Learning Rate and Neighborhood Function of Self-organizing Map (SOM) for Specific Humidity Pattern Classification over Southern Thailand , 2016 .

[40]  Frédéric Alexandre,et al.  Self-organizing Map Initialization , 2005, ICANN.

[41]  Roberto Oberti,et al.  Plant disease detection based on data fusion of hyper-spectral and multi-spectral fluorescence imaging using Kohonen maps , 2005, Real Time Imaging.

[42]  Sildomar T. Monteiro,et al.  Machine learning based hyperspectral image analysis: A survey , 2018, ArXiv.

[43]  Helge J. Ritter,et al.  Neural computation and self-organizing maps - an introduction , 1992, Computation and neural systems series.

[44]  Eric Jones,et al.  SciPy: Open Source Scientific Tools for Python , 2001 .

[45]  Philip J. Howarth,et al.  Hyperspectral remote sensing for estimating biophysical parameters of forest ecosystems , 1999 .

[46]  Teuvo Kohonen,et al.  Self-organized formation of topologically correct feature maps , 2004, Biological Cybernetics.

[47]  C. Ji,et al.  Land-Use Classification of Remotely Sensed Data Using Kohonen Self-organizing Feature Map Neural Networks , 2006 .

[48]  Teuvo Kohonen,et al.  Essentials of the self-organizing map , 2013, Neural Networks.

[49]  Victor Sousa Lobo,et al.  Application of Self-Organizing Maps to the Maritime Environment , 2009, IF&GIS.