Abstract
This paper presents a time series whole clustering system that incrementally constructs a tree-like hierarchy of clusters, using a top-down strategy. The Online Divisive-Agglomerative Clustering (ODAC) system uses a correlation-based dissimilarity measure between time series over a data stream and possesses an agglomerative phase to enhance a dynamic behavior capable of concept drift detection. Main features include splitting and agglomerative criteria based on the diameters of existing clusters and supported by a. significance level. At each new example, only the leaves are updated, reducing computation of unneeded dissimilarities and speeding up the process every time the structure grows. Experimental results on artificial and real data suggest competitive performance on clustering time series and show that the system is equivalent to a batch divisive clustering on stationary time series, being also capable of dealing with concept drift. With this work, we assure the possibility and importance of hierarchical incremental time series whole clustering in the data stream paradigm, presenting a. valuable and usable option.

Abstract
This paper reports new results in adopting entropy concepts to the training of mappers such as neural networks to perform wind power prediction as a function of wind characteristics (mainly speed and direction) in wind parks connected to a power grid. It also addresses the differences relevant to power system operation between off-line and on-line training of neural networks. Real case examples are presented.

Abstract
This paper reports new results in adopting entropy concepts to the training of neural networks to perform wind power prediction as a function of wind characteristics (speed and direction) in wind parks connected to a power grid. Renyi's entropy is combined with a Parzen windows estimation of the error pdf to form the basis of two criteria (minimum entropy and maximum correntropy) under which neural networks are trained. The results are favorably compared in online and offline training with the traditional minimum square error (MSE) criterion. Real case examples for two distinct wind parks are presented.

Abstract
The motivation for this work is the study and prediction of wind ramp events occurring in a large-scale wind farm located in the US Midwest. In this paper we introduce the SHRED framework, a stream-based model that continuously learns a discrete HMM model from wind power and wind speed measurements. We use a supervised learning algorithm to learn HMM parameters from discretized data, where ramp events are HMM states and discretized wind speed data are HMM observations. The discretization of the historical data is obtained by running the SAX algorithm over the first order variations in the original signal. SHRED updates the HMM using the most recent historical data and includes a forgetting mechanism to model natural time dependence in wind patterns. To forecast ramp events we use recent wind speed forecasts and the Viterbi algorithm, that incrementally finds the most probable ramp event to occur. We compare SHRED framework against Persistence baseline in predicting ramp events occurring in short-time horizons, ranging from 30 minutes to 90 minutes. SHRED consistently exhibits more accurate and cost-effective results than the baseline. © 2012 Springer-Verlag Berlin Heidelberg.

Abstract
This paper reports new results in adopting entropy concepts to the training of mappers such as neural networks to perform wind power prediction as a function of wind characteristics (mainly speed and direction) in wind parks connected to a power grid. Renyi's Entropy is combined with a Parzen Windows estimation of the error pdf to form the basis of three criteria (MEE, MCC and MEEF) under which neural networks are trained. The results are favourably compared with the traditional minimum square error (MSE) criterion. Real case examples for two distinct wind parks are presented.

Abstract