Humans postural sways: non-Gauss, variability, and aging aspects of a data set mining

Abstract

This report presents the results of data mining of a sample with two focus-groups. They have the same sizes: seven people into each one, random-wise picked up from the same data set. The trials of groups were idem. The postural swings, that is, the move of the center-of-pressure (COP), were recorded. Maple, a computer math system, has allowed us to apply the Principal Components Analysis, Statistical analysis, Kernel Density Estimations (KDE) for the probabilities. Poincare and Recurrence Plots were other tools for modern data mining. The non-Gauss features of the real distributions are not that to be neglect. They exist not only as outliers but as sharp kurtosis and skewness. Still, they are so far not enow to grave doubts to the Fractional Brownian model. We found some subtle aging hallmarks for focus-groups. First, the trend of variability descriptors to be bi-modal is sheerer for the older group. Second, the coefficients of correlations of the short-time variability index with other descriptors are clear age-related.2012 ACM Subject ClassificationApplied computing → Life and medical sciences → Health informatics

Обробка Даних Про Рецидиви Раку Молочної Залози Для Більш Надійного Прогнозу

“…The authors would like to underline two limitations of the upgraded dataset that is accessible in [13]: First, the dataset still holds 16 instances from 252, with two missed attributes from six; perhaps these instances should also be removed.…”

“…To compare the performance of the two datasets, we analyzed the main performance indexes of the original [4] and upgraded [13] datasets using a version of the J48 classifier. Specifically, we used "weka.classifiers.trees.J48Consolidated -A -C 0.25 -M 2 -Q 1 -RM-C -RM-N 99.0 -RM-B -2 -RM-D 50.0".…”

“…The dataset [13] is helpful in machine learning models' devising, which shall classify, detect, and forecast probabilities of recurrence events of breast cancer in clinics.…”