Author(s): Yamuna Devi S Dr. K. Sasikala Rani2 M.E.,Phd2, Pavithra M3, Priyanga M4
For more than a decade, time series similarity search has been given a great deal of attention by data mining researchers. As a result, many time series representations and distance measures have been proposed. The works on time series similarity search relies on shape-based similarity matching. Some of the algorithms are potential for short time series data and some of the data mining approaches are more powerful for long time series data. In existing scenario, the method is introduced named as Collective of Transformation-Based Ensembles method (COTE). It is mainly used for increasing the classification accuracy than preceding research. Another algorithm is named as Time series classification (TSC) which is used for transformation process which is based on comparative features. COTE contains classifiers constructed in the time, frequency, change, and shapelet transformation domains combined in alternative ensemble structures. However it has issue with transformation process and hence accuracy of the classification is reduced significantly. To avoid this issue we go for proposed scenario. In proposed system, we introduced the concept called as run length transformation to improve the classification accuracy higher than existing system. The run length algorithm is improved along with genetic approach to produce the optimal features. In this scenario, the measures are considered as similarity coefficient, likelihood ratio and dynamic time warping (DTW). Based on the modified k- nearest neighbor distance concept the speed is increased and classification accuracy is improved prominently. From the experimental result we can conclude that our proposed scenario yields better classification performance rather than existing scenario.