TY - JOUR

T1 - Incorporating spatial association into statistical classifiers

T2 - local pattern-based prior tuning

AU - Bai, H.

AU - Cao, F.

AU - Atkinson, M.P.

AU - Chen, Q.

AU - Wang, J.

AU - Ge, Y.

N1 - This is an Accepted Manuscript of an article published by Taylor & Francis in International Journal of Geographical Information Science on 10/03/2020, available online: https://www.tandfonline.com/doi/full/10.1080/13658816.2020.1737702

PY - 2020/9/1

Y1 - 2020/9/1

N2 - This paper proposes a new classification method for spatial data by adjusting prior class probabilities according to local spatial patterns. First, the proposed method uses a classical statistical classifier to model training data. Second, the prior class probabilities are estimated according to the local spatial pattern and the classifier for each unseen object is adapted using the estimated prior probability. Finally, each unseen object is classified using its adapted classifier. Because the new method can be coupled with both generative and discriminant statistical classifiers, it performs generally more accurately than other methods for a variety of different spatial datasets. Experimental results show that this method has a lower prediction error than statistical classifiers that take no spatial information into account. Moreover, in the experiments, the new method also outperforms spatial auto-logistic regression and Markov random field-based methods when an appropriate estimate of local prior class distribution is used.

AB - This paper proposes a new classification method for spatial data by adjusting prior class probabilities according to local spatial patterns. First, the proposed method uses a classical statistical classifier to model training data. Second, the prior class probabilities are estimated according to the local spatial pattern and the classifier for each unseen object is adapted using the estimated prior probability. Finally, each unseen object is classified using its adapted classifier. Because the new method can be coupled with both generative and discriminant statistical classifiers, it performs generally more accurately than other methods for a variety of different spatial datasets. Experimental results show that this method has a lower prediction error than statistical classifiers that take no spatial information into account. Moreover, in the experiments, the new method also outperforms spatial auto-logistic regression and Markov random field-based methods when an appropriate estimate of local prior class distribution is used.

KW - spatial auto-logistic regression

KW - spatial data

KW - Spatial pattern

KW - statistical classifier

U2 - 10.1080/13658816.2020.1737702

DO - 10.1080/13658816.2020.1737702

M3 - Journal article

VL - 34

SP - 2077

EP - 2114

JO - International Journal of Geographical Information Science

JF - International Journal of Geographical Information Science

SN - 1365-8816

IS - 10

ER -