Energy Time Series Forecasting Based on Pattern Sequence Similarity

Abstract

Abstract-This paper presents a new approach to forecast the behavior of time series based on similarity of pattern sequences. First, clustering techniques are used with the aim of grouping and labeling the samples from a data set. Thus, the prediction of a data point is provided as follows: first, the pattern sequence prior to the day to be predicted is extracted. Then, this sequence is searched in the historical data and the prediction is calculated by averaging all the samples immediately after the matched sequence. The main novelty is that only the labels associated with each pattern are considered to forecast the future behavior of the time series, avoiding the use of real values of the time series until the last step of the prediction process. Results from several energy time series are reported and the performance of the proposed method is compared to that of recently published techniques showing a remarkable improvement in the prediction.

A New Methodology Based on Imbalanced Classification for Predicting Outliers in Electricity Demand Time Series

“…A method for the prediction of outlier occurrence was proposed in [12]. In particular, The Pattern Sequence Forecasting (PSF) algorithm [13] was adapted to deal with spike values in the field of electricity price forecasting. As a case study, the markets of New York, Australia, and the Iberian Peninsula were examined.…”

An evolutionary algorithm to discover quantitative association rules in multidimensional time series

“…Usually, the sliding window concept has been successfully applied to forecast time series (Martínez-Á lvarez et al 2011;Nikolaidou and Mitkas 2009). However, this concept has been recently used in (Khan et al 2010) with the purpose of obtaining a low use of memory and low proposed the use of some new operators such as rounding, repairing or filtrating.…”

Clustering and Support Vector Regression for Water Demand Forecasting and Anomaly Detection

“…With respect to the idea of performing the two consecutive stages of analysis proposed in this paper, a similar solution was suggested to forecast energy consumption: in that case, the forecasting procedure was driven by a preliminary clustering of pattern sequences (i.e., energy consumption time-series) [29], and more recently, a general algorithm for pattern sequence-based forecasting has been released [30]. The main difference is in generating the forecasts: in [29,30], the overall data stream is transformed into a sequence of cluster assignments with a cluster label for each day; the forecast is computed by averaging the daily consumption patterns, which occur after a given sub-sequence of cluster labels, while the approach proposed in this paper uses the results of clustering to perform a supervised learning stage inferring a number of forecasting models-one for each cluster and for each hourly consumption to be predicted.…”

“…The main difference is in generating the forecasts: in [29,30], the overall data stream is transformed into a sequence of cluster assignments with a cluster label for each day; the forecast is computed by averaging the daily consumption patterns, which occur after a given sub-sequence of cluster labels, while the approach proposed in this paper uses the results of clustering to perform a supervised learning stage inferring a number of forecasting models-one for each cluster and for each hourly consumption to be predicted.…”