Abstract: It is shown that the classical Chi-Square test has insufficient capacity for efficient processing of biometric data. It is shown that there is a possibility to increase the power of statistical processing through the use of several well-known statistical tests, through the neural network combining their private decisions. Contains tables of formulas promising statistical criteria that complement already used statistical tests. Considered the influence of quantization errors caused by the small amount of experience in the test sample. Proposed to raise the reliability of the estimates due to the digital smoothing of histograms with uniform quantization step. Shows the tables and nomograms to assess the reduction in the probability of errors of the first and second order transition to multivariate statistical analysis of biometric data.
Berik Akhmetov, Alexander Ivanov, Alexander Malygin, Zhibek Alibiyeva, Kaiyrkhan Mukapil, Gulzhanat Beketova and Nazym Zhumangalieva, 2015. Prospects for Multiple Reductions in Test Samples with a Multivariate, Multicriteria, The Neural Network Statistical Analysis of Biometric Data. Research Journal of Applied Sciences, 10: 956-967.