Problems of Tribology = Проблеми трибології - 2015 рік
Постійне посилання зібрання
Переглянути
Перегляд Problems of Tribology = Проблеми трибології - 2015 рік за Ключові слова "539.375.6+539.538+519.237.8"
Зараз показуємо 1 - 3 з 3
Результатів на сторінці
Налаштування сортування
Документ A method of resume-training of discontinuous wear state trackers for composing boosting high-accurate ensembles needed to regard statistical data inaccuracies and shifts(Хмельницький національний університет, 2015) Romanuke, V.V.; Романюк, В.В.For tracking metal wear states at bad statistical data inaccuracies and shifts, there is a method of resume-training of discontinuous wear state trackers for boosting them within high-accurate ensembles. These trackers are Gaussian-noiseddata- trained two-layer perceptrons. An ordinary tracker is selected and, if its performance is satisfactory, it is resumed-trained cyclically. Number of additional passes of training sets is limited. The resume-training procedure wholly can be cycled.Документ Equally - weighted compositions of gaussian - noised - data – trained two - layer perceptrons in boosting ensembles for high - accurate discontinuous tracking of wear states regarding statistical data inaccuracies and shifts(Khmelnitskiy National University, 2015) Romanuke, V.V.; Романюк, В.В.Equally - weighted compositions of Gaussian - noised-data - trained two - layer perceptrons are studied in order to track metal wear states more accurately at the highest level of statistical data inaccuracies and shifts (noise). The noise range is modeled through four magnitudes characterizing ultimate jitters and shifts in wear influencing factors. Accuracy and variance gains of equally - weighted compositions seem to be increasing when noise intensities become lower. When boosting ensembles are composed from ordinary classifiers, high-accurate tracking fails. Only composing ensembles from a lot of the best optimized perceptrons, the accuracy improves by 1,5 % for the averaged tracking error rate and by 7,7 % for the tracking error rate at noise maximum. Here, the boosting appears to have its limit. But ensembles of equally-weighted compositions of perceptrons perform even better than ensembles of perceptrons weighted after training. And for ensuring high-accurate discontinuous tracking of wear states, we just need perceptrons trained by quite different backpropagation methods.Документ Optimizing parameters of the two-layer perceptrons’ boosting ensemble training for accuracy improvement in wear state discontinuous tracking model regarding statistical data inaccuracies and shifts(Khmelnitskiy National University, 2015) Romanuke, V.V.; Романюк, В.В.There is a trial of optimization for improving accuracy in tracking metal tool wear states discontinuously, when the states’ finite set has been statistically tied to the set of representative wear influencing factors. Range of wear states is presumed to be wholly sampled into those factors. The tracker is a static model based on boosting ensemble of two-layer perceptrons with nonlinear transfer functions. It successfully regards statistical data inaccuracies and shifts in a problem of tracking 24 wear states featured with 16 wear influencing factors. Having increased number of classifiers within the ensemble up to 30, the averaged gain with the optimized ensemble is about 56 % in respect of the best ensemble of three classifiers. Similarly, variance of tracking error rate over 24 wear states is about 53 % lower. Nearly the same results are registered when the ensemble is composed without training, but just setting every classifier’s weight to one thirtieth. To get the perfected accuracy more, such equally-weighted compositions shall be investigated in the sequel.