Abstract
Clustering is an important method in hydrological sequence data mining, where dimension deduction is the key efficiency. In this paper, the Mallat algorithm and Daubechies wavelet are used to conduct wavelet transform on hydrological sequences. Through k-level wavelet transform, the hydrological sequences are divided into approximate part A and detailed part B with different time scales. The sequence length at the k-th level is 1/2k times of the length of the original sequence. Thus the efficiency of the algorithm is improved. Real tests show that obtained clusters conform to the real situation.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have