Seasonal Rainfall Prediction Using the Matrix Decomposition Method


Hideo Hirose, Junaida Binti Sulaiman, Masakazu Tokunaga


14th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD 2013) July 1-3, 2013, Honolulu, Hawaii. Studies in Computational Intelligence (R. Lee (Edi)), Volume 492, pp.173-185, DOI: 10.1007/978-3-319-00738-0, Springer 2013.

The matrix decomposition is one of the most powerful methods in recommendation systems. In the recommendation system,
even if evaluation values in a matrix where users and items are corresponding to row and column are provided incompletely,
we can predict the vacant elements of the matrix using the observed values. This method is applied to a variety of the fields, e.g., for movie recommendations, music recommendations, book recommendations, etc. In this paper, we apply the matrix decomposition method to predict the amount of seasonal rainfalls. Applying the method to the case of Indian rainfall data from 1871 to 2011, we have found that the early detection and prediction for the extreme-value of the monthly rainfall can be attained.
Using the newly introduced accuracy evaluation criterion, risky, we can see that the matrix decomposition method using cylinder-type matrix provides the comparative accuracy to the artificial neural network result which has been conventionally used.

Key Words
matrix decomposition; recommendation system; rainfall prediction; early detection; risky; artificial neural networks; time series;



Times Cited in Web of Science:

Cited in Books: