Electricity Forecasting Using Pattern Matching Technique

MSRDG International Journal of Computer Scientific Technology & Electronics Engineering

 

© 2025 by MSRDG IJCSTEE Journal

Volume 1 Issue 1

Year of Publication: 2025



Authors: B. Senthilnayaki, K. Venkatalakshmi
Paper


Download


Article ID
MSRDG-IJCSTEE-V1I1P101
Abstract:

Introduces a new method for time series forecasting based on a pattern matching technique. The procedure involves two steps. Initially, clustering methods are used to partition and tag the dataset, making it easier to represent structured data. Subsequently, a pattern matching forecasting algorithm is utilized to find similar patterns in past data. Predictions are only made where corresponding patterns exist, using labels rather than actual values for increased accuracy. The method has the goal of optimizing forecasting performance through the use of past pattern identification.

Keywords: Pattern sequence, Prediction, Time series, Indexing, Clustering