Abstract
Image processing of particulate phases can provide an important assistance for the estimation of particle size and shape distributions in multiphase systems. The knowledge of these distributions is of major importance in the modeling of agitated liquid-liquid systems either for hydrodynamic and mass transfer (with or without chemical reaction) simulation. Often, obtaining these distributions implies visual/manual techniques for identifying, counting and measuring the particles. This implies high costs, intensive labor, weariness build-up and consequent high error rates. A largely automated computational approach presents a great potential for better performance. In this paper we describe our technique in image processing for shape discrimination and size classification for liquid drops in monochromatic digitized frames. We empirically evaluate our approach on examples of images.

Abstract
In this paper we describe a parameter optimization approach to a simulation algorithm of a mixer-settler system in the transient state. The model we are using for the shallow-layer settler, in a mixer-settler system, is able to describe the hydrodynamic phenomena of the transient state of a liquid liquid system. Its mathematical model includes parameters of the drop transport process as well as of the drop drop and drop-interface coalescence with the active interface. The most adequate values of these parameters are unknown. In order to tune the model parameters we have Linked the mixer-settler simulation algorithm to an optimization procedure. We have used the Hooke-Jeeves optimization algorithm to fit these parameters to given experimental results.

Abstract
In this paper we address the accurate estimation of the frequency of sinusoids of natural signals such as singing, voice or music. These signals are intrinsicly harmonic and are normally contaminated by noise. Taking the Cramér-Rao Lower Bound for unbiased frequency estimators as a reference, we compare the performance of several DFT-based frequency estimators that are non-iterative and that use the rectangular window or the Hanning window. Tests conditions simulate harmonic interference and two new ArcTan-based frequency estimators are also included in the tests. Conclusions are presented on the relative performance of the different frequency estimators as a function of the SNR. ©2010 IEEE.

Abstract
The accurate estimation of the frequency of sinusoids is a frequent problem in many signal processing problems including the real-time analysis of the singing voice. In this paper we rely on a single DFT magnitude spectrum in order to perform frequency estimation in a non-iterative way. Two new frequency estimation methods are derived that are matched to the time analysis window and that reduce the maximum absolute estimation error to about 0.1% of the bin width of the DFT. The performance of these methods is evaluated including the parabolic method as a reference, and considering the influence of noise. A combined model is proposed that offers higher noise robustness than that of a single model. ©2010 IEEE.

Abstract
The ensembling of classifiers tends to improve predictive accuracy. To obtain an ensemble with N classifiers, one typically needs to run N learning processes. In this paper we introduce and explore Model Jittering Ensembling, where one single model is perturbed in order to obtain variants that can be used as an ensemble. We use as base classifiers sets of classification association rules. The two methods of jittering ensembling we propose are Iterative Reordering Ensembling (IRE) and Post Bagging (PB). Both methods start by learning one rule set over a single run, and then produce multiple rule sets without relearning. Empirical results on 36 data sets are positive and show that both strategies tend to reduce error with respect to the single model association rule classifier. A bias-variance analysis reveals that while both IRE and PB are able to reduce the variance component of the error, IRE is particularly effective in reducing the bias component. We show that Model Jittering Ensembling can represent a very good speed-up w.r.t. multiple model learning ensembling. We also compare Model Jittering with various state of the art classifiers in terms of predictive accuracy and computational efficiency.

Abstract
The current quality control methodology adopted by the water distribution service provider in the metropolitan region of Porto - Portugal, is based on simple heuristics and empirical knowledge. Based on the domain complexity and data volume, this application is a perfect candidate to apply data mining process. In this paper, we propose a new methodology to predict the range of normality for the values of different water quality parameters. These intervals of normality are of key importance to decide on costly inspection activities. Our experimental evaluation confirms that our proposal achieves good results on the task of forecasting the normal distribution of values for the following 30 days. The proposed method can be applied to other domains with similar network monitoring objectives.