Research Article
A Hybrid Algorithm for Clustering of Time Series Data Based on Affinity Search Technique
Pseudocode 1
Pseudocode related to TTC.
Method: (, , ) | Input: : the set of time-series | : Affinity threshold | : cluster number | Output: : set of clusters | /* 1. time-series data reduction*/ | (1) -norm() /*-normalization of all time-series */ | (2) ←SmilarityED () /* Calculate disimilarity array based on (5)*/ | (3) (, ← CAST(, ) /* is determined automatically by CAST*/ | (4) for to ; is the number of sub-clusters | (5) ← Average (, ) /* summarize the cluster */ | (6) ← /*: a collection of prototypes */ | (7) end for | /*2. clustering*/ | (8) ←SmilarityDTW() | (9) ← -medoids() | (10) ← labels()validate the MTC clustering results | (11) return |
|