A binary classification problem is imbalanced when the number of samples from the two groups differs. For the high dimensional case, where the number of variables ismuch larger than the number of samples, imbalance leads to a bias in the classification. The independence classifier is studied theoretically and based on the analysis two new classifiers are suggested that can handle any imbalance ...

Purpose: This work focuses on the reduction of scan time required by the phase-contrast (PC) MRA technique. The presented work builds on earlier work [1] in the way of parameter optimization. The proposed method consists of a 3D variable density spiral projection imaging (SPI) trajectory [2] combined with a dual velocity encoding technique [3]. SPI is a rapid imaging technique that improves acq...

One of the particular characteristics of text classification tasks is that they present large class imbalances. Such a problem can easily be tackled using resampling methods. However, although these approaches are very simple to implement, tuning them most effectively is not an easy task. In particular, it is unclear whether oversampling is more effective than undersampling and which oversampli...

This paper proposes a real-time hybrid solution between standard shadow mapping and any of the current perspective shadow mapping based methods. Existing methods are known to have undersampling problems. At little extra cost we combine two methods to minimize undersampling errors. This is done by computing a plane equation which separates the two resulting shadow maps to get the best possible a...

Forecasting the creditworthiness of customers is a central issue banking activity. This task requires analysis large datasets with many variables, for which machine learning algorithms and feature selection techniques are crucial tool. Moreover, percentages “good” “bad” typically imbalanced such that over- undersampling should be employed. In literature, most investigations tackle these three i...

We investigate whether there is a within-speaker effect of a higher F0 on the values of the first and the second formant. When asked to speak at a high F0, speakers turn out to raise their formants as well. In the F1 dimension this effect is greater for women than for men. We conclude that the general formant raising effect might be due to the physiology of a high F0 (i.e. raised larynx and sho...

This paper proposes an economic method for the nonlinear modeling of dynamic processes using feedback neural networks, by undersampling the training sequences. The undersampling (i) allows a better exploration of the operating range of the process for a given size of the training sequences, and (ii) it speeds up the training of the feedback networks. This method is successfully applied to the t...

____________________________________________________________________________ Abstract The present ISR describes the encircled energy(EE) analysis of pinhole images made on the M10 WFC3 1R-MUX at the Goddard Space Flight Center (GSFC) Detector Characterization Lab (DCL) on 29 May 2001. Analysis shows that the pinhole image’s centroid is stable within a 1-pixel area and the 80% of the encircled e...