TY - GEN
T1 - Multisource data classification using a hybrid semi-supervised learning scheme
AU - Vatsavai, Ranga Raju
AU - Badhuri, Budhendra
AU - Shekhar, Shashi
AU - Burk, Thomas E.
PY - 2008
Y1 - 2008
N2 - In many practical situations thematic classes can not be discriminated by spectral measurements alone. Often one needs additional features such as population density, road density, wetlands, elevation, soil types, etc. which are discrete attributes. On the other hand remote sensing image features are continuous attributes. Finding a suitable statistical model and estimation of parameters is a challenging task in multisource (e.g., discrete and continuous attributes) data classification. In this paper we present a semi-supervised learning method by assuming that the samples were generated by a mixture model, where each component could be either a continuous or discrete distribution. Overall classification accuracy of the proposed method is improved by 12% in our initial experiments.
AB - In many practical situations thematic classes can not be discriminated by spectral measurements alone. Often one needs additional features such as population density, road density, wetlands, elevation, soil types, etc. which are discrete attributes. On the other hand remote sensing image features are continuous attributes. Finding a suitable statistical model and estimation of parameters is a challenging task in multisource (e.g., discrete and continuous attributes) data classification. In this paper we present a semi-supervised learning method by assuming that the samples were generated by a mixture model, where each component could be either a continuous or discrete distribution. Overall classification accuracy of the proposed method is improved by 12% in our initial experiments.
KW - Expectation maximization
KW - GMM
KW - Multisource data
KW - Semi-supervised learning
UR - http://www.scopus.com/inward/record.url?scp=67649800351&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=67649800351&partnerID=8YFLogxK
U2 - 10.1109/IGARSS.2008.4779525
DO - 10.1109/IGARSS.2008.4779525
M3 - Conference contribution
AN - SCOPUS:67649800351
SN - 9781424428083
T3 - International Geoscience and Remote Sensing Symposium (IGARSS)
SP - III1016-III1019
BT - 2008 IEEE International Geoscience and Remote Sensing Symposium - Proceedings
T2 - 2008 IEEE International Geoscience and Remote Sensing Symposium - Proceedings
Y2 - 6 July 2008 through 11 July 2008
ER -