Conditions Of Insufficient Data Research Articles

Research on rolling bearing fault diagnosis method based on simulation and experiment fusion drive.

Aiming at the problem that the rolling bearing fault data are difficult to obtain and that the traditional fault diagnosis method does not consider the signal uncertainty characteristics and the low accuracy of models in the process of rolling bearing fault, a fault diagnosis method based on simulation and experiment fusion drive is proposed. First, the dynamics simulation model of rolling bearings under different fault conditions is established to obtain the bearing fault simulation signals. Second, a sequence generative adversarial network is used to fuse the simulation and experimental data. Bearing vibration signals are often very uncertain, so considering the probability characteristics of fault signals, the probability box model under different fault states is constructed by the direct probability box modeling method, and its characteristic vectors are extracted. Finally, an extreme gradient boosting Tree model for fault diagnosis classification is constructed to compare and evaluate the classification and diagnosis effects of bearing states before and after data fusion. The results show that the proposed method has a good diagnostic effect and is suitable for solving the fault diagnosis problem under the condition of insufficient data.

УРОВЕННЫЙ РЕЖИМ АРКТИЧЕСКИХ РЕК ПО ДАННЫМ МОДЕЛИРОВАНИЯ И СПУТНИКОВЫХ ИЗМЕРЕНИЙ

Increasing accuracy of satellite altimetry water level retrievals, especially over rivers with seasonal ice cover, motivated the application of satellite observations for improving hydrodynamic models. Altimetric water level measurements are particularly valuable in areas with limited in situ data on river bed and floodplain topography as well as under conditions of insufficient data on water discharge and water level. The MIKE11 one-dimensional hydrodynamic model was adapted for the Lower Kolyma. The model was calibrated using the water level measurements at the Kolymskoe water gauge station located in the upper part of the reach. The verification of the simulated water levels against altimetry data was performed. The comparison of simulated and satellite-derived water levels at seven locations has shown that the model reproduces well the water level variability throughout the analyzed reach (167 km). The accuracy of the water level simulations as compared to the altimetric measurements varied from 0.20 to 2.75 m (or 13-62% of the seasonal amplitude).

Entity Relationship Extraction Based on a Multi-Neural Network Cooperation Model

Entity relation extraction mainly extracts relations from text, which is one of the important tasks of natural language processing. At present, some special fields have insufficient data; for example, agriculture, the metallurgical industry, etc. There is a lack of an effective model for entity relationship recognition under the condition of insufficient data. Inspired by this, we constructed a suitable small balanced data set and proposed a multi-neural network collaborative model (RBF, Roberta–Bidirectional Gated Recurrent Unit–Fully Connected). In addition, we also optimized the proposed model. This model uses the Roberta model as the coding layer, which is used to extract the word-level features of the text. This model uses BiGRU (Bidirectional Gated Recurrent Unit)–FC (Fully Connected) as the decoding layer, which is used to obtain the optimal relationship of the text. To further improve the effect, the input layer is optimized by feature fusion, and the learning rate is optimized by the cosine annealing algorithm. The experimental results show that, using the small balanced data set, the F1 value of the RBF model proposed in the paper is 25.9% higher than the traditional Word2vec–BiGRU–FC model. It is 18.6% higher than the recent Bert–BiLSTM (Bidirectional Long Short Term Memory)–FC model. The experimental results show that our model is effective.

Swin Unet3D: a three-dimensional medical image segmentation network combining vision transformer and convolution

BackgroundSemantic segmentation of brain tumors plays a critical role in clinical treatment, especially for three-dimensional (3D) magnetic resonance imaging, which is often used in clinical practice. Automatic segmentation of the 3D structure of brain tumors can quickly help physicians understand the properties of tumors, such as the shape and size, thus improving the efficiency of preoperative planning and the odds of successful surgery. In past decades, 3D convolutional neural networks (CNNs) have dominated automatic segmentation methods for 3D medical images, and these network structures have achieved good results. However, to reduce the number of neural network parameters, practitioners ensure that the size of convolutional kernels in 3D convolutional operations generally does not exceed 7 times 7 times 7, which also leads to CNNs showing limitations in learning long-distance dependent information. Vision Transformer (ViT) is very good at learning long-distance dependent information in images, but it suffers from the problems of many parameters. What’s worse, the ViT cannot learn local dependency information in the previous layers under the condition of insufficient data. However, in the image segmentation task, being able to learn this local dependency information in the previous layers makes a big impact on the performance of the model.MethodsThis paper proposes the Swin Unet3D model, which represents voxel segmentation on medical images as a sequence-to-sequence prediction. The feature extraction sub-module in the model is designed as a parallel structure of Convolution and ViT so that all layers of the model are able to adequately learn both global and local dependency information in the image.ResultsOn the validation dataset of Brats2021, our proposed model achieves dice coefficients of 0.840, 0.874, and 0.911 on the ET channel, TC channel, and WT channel, respectively. On the validation dataset of Brats2018, our model achieves dice coefficients of 0.716, 0.761, and 0.874 on the corresponding channels, respectively.ConclusionWe propose a new segmentation model that combines the advantages of Vision Transformer and Convolution and achieves a better balance between the number of model parameters and segmentation accuracy. The code can be found at https://github.com/1152545264/SwinUnet3D.

On the calculation of the average annual flow of the Prokhladnaya River

In order to develop measures for the rational use and protection of water resources, information on the characteristics of water bodies is needed. The flow of water in the river is the most important hydrological characteristic, which is used in various calculations. In practice activities, the determination of the calculated hydrological characteristics is carried out quite often in conditions of insufficient data from hydrometric observations or in their complete absence. This article presents the results of restoring a number of average annual discharges of the Prokhladnaya River, the catchment area of which occupies the southwestern part of the Kaliningrad region. The Instruch River was chosen as an analogue river, which meets the requirements set forth in the regulatory documents. The obtained data on the average annual discharges of the Prokhladnaya River are of great importance, since the Prokhladnaya River basin plays an important role in the water management complex of the region.

An xception model based on residual attention mechanism for the classification of benign and malignant gastric ulcers

To explore the application value of convolutional neural network combined with residual attention mechanism and Xception model for automatic classification of benign and malignant gastric ulcer lesions in common digestive endoscopy images under the condition of insufficient data. For the problems of uneven illumination and low resolution of endoscopic images, the original image is preprocessed by Sobel operator, etc. The algorithm model is implemented by Pytorch, and the preprocessed image is used as input data. The model is based on convolutional neural network for automatic classification and diagnosis of benign and malignant gastric ulcer lesions in small number of digestive endoscopy images. The accuracy, F1 score, sensitivity, specificity and precision of the Xception model improved by the residual attention module for the diagnosis of benign and malignant gastric ulcer lesions were 81.411%, 81.815%, 83.751%, 76.827% and 80.111%, respectively. The superposition of residual attention modules can effectively improve the feature learning ability of the model. The pretreatment of digestive endoscopy can remove the interference information on the digestive endoscopic image data extracted from the database, which is beneficial to the training of the model. The residual attention mechanism can effectively improve the classification effect of Xception convolutional neural network on benign and malignant lesions of gastric ulcer on common digestive endoscopic images.

КОНЦЕПЦІЯ МІСЬКОГО МЕТАБОЛІЗМУ ЯК ПІДҐРУНТЯ ДЛЯ РОЗБУДОВИ ЦИРКУЛЯРНИХ МІСТ В УКРАЇНІ

The paper goal is to investigate the sustainability of Ukrainian cities from the perspective of urban metabolism model. Based on the study of the features and advantages of the concept of urban metabolism for the development of sustainable and circular cities, the authors applied and analyzed the integral indicators of resource consumption, waste production, economic functions of cities, and also analyzed the institutional capacities, plans and environmental policy measures of 14 regional centers of Ukraine. The scoring helps to determine the ratio of the urban metabolism indicators of the selected cities of Ukraine and to identify the main imbalances in the consumption and use of resources, as well as to identify the cities that require the activation of institutional support for sustainable development. The application of such method highlighted that the urban metabolism concept works and can be applied in conditions of insufficient data and statistical information, as it is now in Ukraine. In research result it is clear that Vinnytsia and Khmelnitskyi are the most sustainable among the selected cities from the perspective of circular economy. Rivne, Kharkiv and Dnipro are characterized by balanced flows, but have higher rates of resource consumption and waste production. Poltava, Odesa, Cherkasy and Sumy require significant environmental policy adjustments in relation to resource consumption and waste management. The capital of Ukraine, Kyiv occupies an average position among all cities, which is enhanced by significant economic and innovative development and high level of public environmental activity. The lowest level of anthropogenic pressure is observed in Chernivtsi.

Mathematical model of bank scoring in conditions of insufficient data

Recently, different methods of object classification using training datasets is actually. One of these methods is naive Bayesian classifier. Class of objects can consist of low number of elements. Such class is called poor class. In this paper we consider classification problem in poor class. Logical classifier doesn’t work in this case. Metric classifier can give good results if and only if there are quite dense set of metrically nearby classified objects in neighborhood of the considering object. Bayesian classifier reevaluates all hypotheses about belonging of the object to certain class. Therefore, Bayesian classifier can solve this classification problem. For example, we considered classic problem of bank scoring. This scoring is based on two criteria. Classified object has two belonging hypotheses. We can apply such reasoning for more difficult cases.

Reed Sternberg Cell Recognition in Hodgkin’s Lymphoma

A technique for the morphological diagnosis of Reed Sternberg cells in Hodgkin’s lymphoma using artificial neural networks is proposed. A flexible architecture of a software complex comprising several modules for automatic biopsy analysis is developed. These software modules, consisting of artificial neural networks, are necessary for processing raw data. Various techniques for training neural networks under conditions of insufficient data for the current and similar problems are analyzed.

Insufficient Data Can Also Rock! Learning to Converse Using Smaller Data with Augmentation

Recent successes of open-domain dialogue generation mainly rely on the advances of deep neural networks. The effectiveness of deep neural network models depends on the amount of training data. As it is laboursome and expensive to acquire a huge amount of data in most scenarios, how to effectively utilize existing data is the crux of this issue. In this paper, we use data augmentation techniques to improve the performance of neural dialogue models on the condition of insufficient data. Specifically, we propose a novel generative model to augment existing data, where the conditional variational autoencoder (CVAE) is employed as the generator to output more training data with diversified expressions. To improve the correlation of each augmented training pair, we design a discriminator with adversarial training to supervise the augmentation process. Moreover, we thoroughly investigate various data augmentation schemes for neural dialogue system with generative models, both GAN and CVAE. Experimental results on two open corpora, Weibo and Twitter, demonstrate the superiority of our proposed data augmentation model.

High-resolution DOA estimation with Meridian prior

Based on the assumption that only a few point sources exist in the spatial spectrum, the direction-of-arrival (DOA) estimation problem can be formulated as a problem of sparse representation of signal with respect to a dictionary. By choosing a proper dictionary, the array measurements can be well approximated by a linear combination of a few entries of the dictionary, in which the non-zero elements of the sparse coefficient vector correspond to the targets’ arrival direction. Conventionally, the desired sparsity of signal is guaranteed by imposing a constraint of Laplace prior on the distribution of signal. However, its performance is not satisfied under the condition of insufficient data or noisy environment since a lot of false targets will appear. Considering that the Meridian distribution has the characteristic of high energy concentration, we propose to adopt the Meridian prior as the prior distribution of the coefficient vector. Further, we present a new minimization problem with the Meridian prior assumption (MMP) for DOA estimation. Because the Meridian prior imposes a more stringent constraint on the energy localization than the Laplace prior, the proposed MMP method can achieve a better DOA estimation, which is embodied in higher resolution and less false targets. The experiments of both simulation and ground truth data process exhibit the superior performance of our proposed algorithm.

An EigenMLLR-Like Eigen-FLS Approach for Speech Pattern Recognition

The Eigen-FLS using an eigenspace-based scheme to build up fuzzy logic system (FLS) fast for speech pattern recognition applications has been developed in the author’s previous works. However, speech pattern recognition by Eigen-FLS will still encounter a dissatisfactory recognition performance when the collected data for eigen value calculations of the FLS eigenspace, i.e. the eigen-decomposition process, is scarce. To regulate the influence of Eigen-FLS when data from a test speaker for eigen-decomposition is improper, this paper proposes an EigenMLLR-like Eigen-FLS approach. The developed EigenMLLR-like Eigen-FLS integrates the kernel idea of EigenMLLR speaker adaptation for properly adjusting the target speaker’s Eigen-FLS model in the eigenspace of FLS. EigenMLLR-like Eigen-FLS developed in this paper will be more robust than conventional Eigen-FLS in a speech pattern recognition application with an adverse condition of insufficient data from the speaker.

Equipment failure rate updating—Bayesian estimation

The paper presents a Bayes’ method for augmenting generic equipment failure data with a prior distribution – predicated on the evidence, e.g., plant data – resulting in a posterior distribution. The depth of the evidence is significant in shaping the characteristics of the posterior distribution. In conditions of insufficient data about the prior distribution or great uncertainty in the generic data sources, we may use “constrained non-informative priors”. This representation of the prior preserves the mean value of the failure rate estimate and maintains a broad uncertainty range to accommodate the site-specific event data. Although the methodology and the case study presented in this paper focus on the calculation of a time-based (i.e., failures per unit time) failure rate, based on a Poisson likelihood function and the conjugate gamma distribution, a similar method applies to the calculation of demand failure rates utilizing the binomial likelihood function and its conjugate beta distribution.

On estimating the hydraulic properties of soil, Part 2. A new empirical equation for estimating hydraulic conductivity for sands

A new empirical equation to estimate hydraulic conductivity is proposed, based on a large set of measured data for hydraulic properties of soil. The equation is simpler and more accurate than the series-parallel model. Under conditions of insufficient data, the new equation provides a good estimation of hydraulic conductivity for sands. For the same class of soils, another empirical equation is proposed to estimate the power N in the Averjanov-Irmay function.

Stages of Development and the uses of Planning: Some Nepali Experiences

In their development efforts, less developed countries (LDCs) often seek to relieve the uncertainty by which they are frequently characterized by adopting the technique of national planning. Such planning has proved successful in technologically advanced nations (such as the USSR), but it does not appear that LDCs have considered carefully whether they are equipped to implement it. In fact, planning among the LDCs has frequently been an exercise of little relevance to actual development operations, even when they have concentrated development efforts on a single sector or in a selected region of the country. Nevertheless, this sustained phenomenon has rarely caused the LDCs to abandon hope in the promise of planning or to cease the preparation of consecutive development plans.1 As LDCs gradually adopt planning, they experience a process that is implicit in the solution that planning is intended to offer and that serves, in practice, as a frame of reference for actual development activity. It is contended here that this process consists of three major stages, each characterized by a different use of planning activities and products (i.e., plans). In the first stage national planning is adopted but turns into a mere ritual that fails to lend itself to actual development activity. During this stage plans are prepared under conditions of insufficient data and acutely limited administrative capacity as well as amidst volatility in both the political system and the bureaucratic machinery. Consequently, not only are the plans unable to serve as a meaningful framework for action, but the circumstances surrounding the development effort undermine the translation of plans into practice. The actual development activities carried out during this stage are decided upon the basis of immediate and rapidly changing political opportunities