Calibration of assessment instruments – classification in mathematics at the Universidad Jorge Tadeo Lozano

Item Response Theory, IRT, is a valuable methodology for analyzing the quality of the instruments utilized in assessment of academic achievement. This article presents an implementation of the mentioned theory, particularly of the Rasch model, in order to calibrate items and the instrument used in the classification test for the Basic Mathematics subject at Universidad Jorge Tadeo Lozano. 509 responses chains of students, obtained in the june 2011 application, were analyzed with a set of 45 items, through eight case studies that are showing progressive steps of calibration. Criteria of validity of items and of whole instrument were defined and utilized, to select groups of responses chains and items that were finally used in the determination of parameters which then allowed the classification of assessed students by the test.

On the origin of the forward peak and backward oscillations in the F + H2(v = 0) to HF(v' = 2) + H reaction

We present a semiclassical complex angular momentum (CAM) analysis of the forward scattering peak which occurs at a translational collision energy around 32 meV in the quantum mechanical calculations for the F + H2(v = 0, j = 0) ? HF(v' = 2, j' = 0) + H reaction on the Stark–Werner potential energy surface. The semiclassical CAM theory is modified to cover the forward and backward scattering angles. The peak is shown to result from constructive/destructive interference of the two Regge states associated with two resonances, one in the transition state region and the other in the exit channel van der Waals well. In addition, we demonstrate that the oscillations in the energy dependence of the backward differential cross section are caused by the interference between the direct backward scattering and the decay of the two resonance complexes returning to the backward direction after one full rotation.

Support vector machine classification for large data sets via minimum enclosing ball clustering

Support vector machine (SVM) is a powerful technique for data classification. Despite of its good theoretic foundations and high classification accuracy, normal SVM is not suitable for classification of large data sets, because the training complexity of SVM is highly dependent on the size of data set. This paper presents a novel SVM classification approach for large data sets by using minimum enclosing ball clustering. After the training data are partitioned by the proposed clustering method, the centers of the clusters are used for the first time SVM classification. Then we use the clusters whose centers are support vectors or those clusters which have different classes to perform the second time SVM classification. In this stage most data are removed. Several experimental results show that the approach proposed in this paper has good classification accuracy compared with classic SVM while the training is significantly faster than several other SVM classifiers.