A novel algorithm applied to classify unbalanced data

Using Monte Carlo method to estimate the behavior of neural training between balanced and unbalanced data in classification of patterns

“…The last inequality equation of (21) is re-presented as follows: (22) So, the inequality (22) can be unraveled in three cases: (23) Values of e where λ ∞ is a great value where a total imbalance of data occurs (it is an adjustment constant) and 1 < n 21 = n2 n1 < λ ∞ . As it can be noticed, there are two allowed interpolations to determine e 2 m1 : one is linear; and the other is exponential.…”

“…Only by mere convention to be followed from beginning to end of this article, the values associated with n 2 will express the sense of abundance of data and the values associated with n 1 will express the sense of data scarcity. There are several papers in the literature (see [19][20][21][22][23] ) describing class imbalance in pattern classification problems. However, as described here, nothing was found about it.…”

“…In study [22] the novel algorithm based on analysis of variance (ANOVA) combined with fuzzy c-means (FCM) and bacterial foraging optimization (BFO) are proposed to classify unbalanced data. The algorithm is supported by simulation results.…”

Two Methods of Automatic Evaluation of Speech Signal Enhancement Recorded in the Open-Air MRI Environment

“…This approach focuses on testing whether there is a common mean of speech features from several groups. Besides the ANOVA F-test giving the ratio of variances between and within groups [16], the hypothesis probability resulting from the Wilcoxon test [25] or the Mann-Whitney U test [26] comparing whether two samples come from identical distributions with equal medians or they do not have equal medians, the Ansari-Bradley hypothesis test [27] is used to specify whether two distributions are the same or they differ in their variances. For a chosen significance level the resulting logical value "0" denotes that the null hypothesis cannot be rejected and the value "1" indicates that it can be rejected.…”

“…The objective approaches for measuring the speech signal quality [14] comprise, for example, evaluation of differences between the speech spectral envelopes [11] or spectral distances [12], etc. These features may be compared and matched using the statistical approaches, like the analysis of variances (ANOVA) [15], [16] or hypothesis tests [17], [18]. The final evaluation in these approaches bears the form of a recognition score that can be obtained by the methods based on artificial neural networks, the nearest neighbor [19], vector quantization classifiers [20], hidden Markov models [21], and support vector machines (SVM) [22].…”