Search results

The electronic nose is an array of chemical or gas sensors and associated with a pattern-recognition framework competent in identifying and classifying odorant or non-odorant and simple or complex gases. Despite more than 30 years of research, the robust e-nose device is still limited. Most of the challenges towards reliable e-nose devices are associated with the non-stationary environment and non-stationary sensor behaviour. Data distribution of sensor array response evolves with time, referred to as non-stationarity. The purpose of this paper is to provide a comprehensive introduction to challenges related to non-stationarity in e-nose design and to review the existing literature from an application, system and algorithm perspective to provide an integrated and practical view.

The authors discuss the non-stationary data in general and the challenges related to the non-stationarity environment in e-nose design or non-stationary sensor behaviour. The challenges are categorised and discussed with the perspective of learning with data obtained from the sensor systems. Later, the e-nose technology is reviewed with the system, application and algorithmic point of view to discuss the current status.

The discussed challenges in e-nose design will be beneficial for researchers, as well as practitioners as it presents a comprehensive view on multiple aspects of non-stationary learning, system, algorithms and applications for e-nose. The paper presents a review of the pattern-recognition techniques, public data sets that are commonly referred to as olfactory research. Generic techniques for learning in the non-stationary environment are also presented. The authors discuss the future direction of research and major open problems related to handling non-stationarity in e-nose design.

The authors first time review the existing literature related to learning with e-nose in a non-stationary environment and existing generic pattern-recognition algorithms for learning in the non-stationary environment to bridge the gap between these two. The authors also present details of publicly available sensor array data sets, which will benefit the upcoming researchers in this field. The authors further emphasise several open problems and future directions, which should be considered to provide efficient solutions that can handle non-stationarity to make e-nose the next everyday device.

The purpose of this paper is to examine neural document clustering techniques, e.g. self‐organising map (SOM) or growing neural gas (GNG), usually assume that textual information is stationary on the quantity.

The authors propose a novel dynamic adaptive self‐organising hybrid (DASH) model, which adapts to time‐event news collections not only to the neural topological structure but also to its main parameters in a non‐stationary environment. Based on features of a time‐event news collection in a non‐stationary environment, they review the main current neural clustering models. The main deficiency is a need of pre‐definition of the thresholds of unit‐growing and unit‐pruning. Thus, the dynamic adaptive self‐organising hybrid (DASH) model is designed for a non‐stationary environment.

The paper compares DASH with SOM and GNG based on an artificial jumping corner data set and a real world Reuters news collection. According to the experimental results, the DASH model is more effective than SOM and GNG for time‐event document clustering.

A real world environment is dynamic. This paper provides an approach to present news clustering in a non‐stationary environment.

Text clustering in a non‐stationary environment is a novel concept. The paper demonstrates DASH, which can deal with a real world data set in a non‐stationary environment.

Using the data set about the restaurants from different countries and their customer's feedback, the purpose of this paper is to address the following issues: in the restaurant industry, how have user behavior and preferences changed during the COVID-19 restrictions period, how did these changes influence the performance of recommendation algorithms and which methods can be proposed to improve the quality of restaurant recommendations in a lockdown scenario.

To assess changes in user behavior and preferences, quantitative and qualitative data analysis was performed to assess the changes in user behavior and preferences. The authors compared the situation before and during the COVID-19 restrictions period. To evaluate the performance of restaurant recommendation systems in a non-stationary setting, the authors tested state-of-the-art collaborative filtering algorithms. This study proposes and investigates a filtering-based approach to improve the quality of recommendation algorithms for a lockdown scenario.

This study revealed that during the COVID-19 restrictions period, the average rating values and the number of reviews have changed. The experimental study confirmed that: the performance of all state-of-the-art recommender systems for the restaurant industry has significantly degraded during the COVID-19 restrictions period; and the accuracy and the stability of restaurant recommendations in non-stationary settings may be improved using the sliding window and post-filtering methods.

The authors propose two novel methods: the sliding window and closed restaurants post-filtering method based on the CatBoost classification model. These methods can be applied to classical collaborative recommender algorithms and increase the value of metrics under non-stationary conditions. These methods can be helpful for developers of recommender systems and massive aggregators of restaurants and hotels. Thus, it benefits both the app end-user and business owners because users honestly rate restaurants when they receive good recommendations and do not downgrade because of external factors.

To the best of the authors’ knowledge, this paper provides the first extensive and multifaceted experimental study of the impact of COVID-19 restrictions on the effectiveness of restaurant recommendation systems in different countries. Two novel methods to tackle restaurant recommendations' performance degradation are proposed and validated.

利用关于不同国家餐厅及其顾客反馈的数据, 我们探索了以下问题：(i) 在餐饮行业, 用户行为和偏好在COVID-19限制期间如何改变, (ii) 这些变化如何影响推荐算法的性能, 以及 (iii) 可以提出哪些方法来改进封锁情景下的餐厅推荐质量。

为了评估用户行为和偏好的变化, 本研究进行了定量和定性数据分析, 对比了COVID-19限制期前后的情况。为了评估非稳态环境中餐厅推荐系统的性能, 我们测试了最先进的协同过滤算法。我们提出并研究了一种基于过滤的方法, 以提高封锁情景下推荐算法的质量。

研究发现, 在COVID-19限制期间, 平均评分和评论数量发生了变化。实验研究证实：(i) 在COVID-19限制期间, 所有最先进的餐厅行业推荐系统的性能显著下降; (ii) 使用滑动窗口和后过滤方法可以改进非稳态环境下餐厅推荐的准确性和稳定性。

我们提出了两种新方法：基于CatBoost分类模型的关闭餐厅后过滤和滑动窗口方法。这些方法可以应用于经典的协同过滤推荐算法, 并在非稳态条件下提高指标值。这些方法对于推荐系统的开发者和大规模餐厅和酒店聚合平台都有帮助。因此, 这对于应用的最终用户和企业主都有好处, 因为当用户得到良好的推荐时, 他们会诚实地对餐厅进行评价, 而不会因为外部因素降低评分。

本文首次提供了COVID-19限制对不同国家餐厅推荐系统有效性影响的广泛多方面的实验研究, 并提出和验证了两种解决餐厅推荐性能下降问题的新方法。

The purpose of this paper is to introduce methods for calculating steady‐state and transient processes in a symmetrical three‐phase matrix‐reactance frequency converter (MRFC). The MRFC in question makes it possible to obtain a load output voltage much greater than the input voltage.

MRFCs based on a matrix‐reactance chopper are used for both frequency and voltage transformation. The processes in a MRFC system are described by nonstationary differential equations. A two‐frequency complex function method is proposed for solving non‐stationary equations in steady‐state. The method is applied to a state‐space averaged mathematical model used in the analysis of the discussed MRFC. A two‐frequency matrix transform is proposed for solving non‐stationary equations. This method can be used to find both transient and steady‐state processes.

The two‐frequency complex function method permits the reduction from 12 non‐stationary differential equations to four stationary differential equations. The two‐frequency matrix transform allows the transformation of non‐stationary differential equations to stationary ones. By using these methods descriptions of steady‐state and transient properties of buck‐boost MRFCs are obtained.

A new method of solving of nonstationary differential equations is presented. The method is useful for process analyses in nonstationary power electronic converters.

Prior studies provide mixed propositions on whether earnings levels or earnings changes provide the better explanatory power for variations of stock returns and whether the time-series behavior of earnings affects the value relevance of both earnings variables. This paper aims to compare the value relevance of earnings levels with that of earnings changes in the return-earnings relations.

The unobservable components model is used to estimate permanent and transitory components of earnings.

The finding shows that the proxy ability of earnings changes for unexpected earnings is sensitive to a firm's time-series earnings permanence property and is unstable and noisy when earnings contain predominantly transitory components, but that of earnings levels is not. The results support earnings levels are a stable and better value relevant proxy in the return-earnings relations.

The findings imply that the valuation role of earnings levels is important in the research relating to earnings components, earnings innovations, and equity valuation, especially when earnings permanence is of interest.

The results provide a new understanding on the role of earnings levels in many business decisions such as executive compensations, institutional investment and conservative accounting where they often involve the choice of using levels and/or changes of earnings variables in making decisions.

The paper contributes to the accounting literature by providing a new insight into the valuation role of earnings levels in the return-earnings relations. The stable value relevance of earnings levels also has important implications, especially for studies that use only earnings levels to assess earnings quality and earnings attributes.

The purpose of this paper is to compare the accuracy of tracking the amplitude and frequency changes of non‐stationary electric signals.

Short‐time fourier transform (STFT) and S‐transform algorithms were applied to analyze non‐stationary signals originating from switching of capacitor banks in a power system.

The S‐transform showed possibilities of sharp localization of the basic component, and allowed improvement of tracking dynamism the transient components in comparison to STFT.

S‐transform is a better tool for the analysis of non‐stationary waveforms in power systems and its properties can be used for diagnostic and power quality applications.

The dynamic tracking of the changes in time and frequency of real‐like signals originating from a power system are investigated in this paper.

This paper aims to explore a new way to extract the fault feature of a rolling bearing signal on the basis of a combinatorial method.

By combining local mean decomposition (LMD) with Teager energy operator, a new feature-extraction method of a rolling bearing fault signal was proposed, called the LMD–Teager transform method. The principles and steps of method are presented, and the physical meaning of the time–frequency power spectrum and marginal spectrum is discussed. On the basis of comparison with the fast Fourier transform method, a simulated non-stationary signal was processed to verify the effect of the new method. Meanwhile, an analysis was conducted by using the recorded vibration signals which include inner race, out race and bearing ball fault signal.

The results show that the proposed method is more suitable for the non-stationary fault signal because the LMD–Teager transform method breaks through the difficulty of the Fourier transform method that can process only the stationary signal. The new method can extract more useful information and can provide better analysis accuracy and resolution compared with the traditional Fourier method.

Combining the advantage of the local mean decomposition and the Teager energy operator, the LMD–Teager method suits the nature of the fault signal. A marginal spectrum obtained from the LMD–Teager method minimizes the estimation bias brought about by the non-stationarity of the fault signal. So, the LMD–Teager transform has better analysis accuracy and resolution than the traditional Fourier method, which provides a good alternative for fault diagnosis of the rolling bearing.

Stimulated by previous reference, which proposed making straight line of regression to test gear gravimetric wear loss sequence distribution, this paper aims to propose using straight line of regression to fit gear gravimetric wear loss sequence based on stationary random process suppose. Faced to that the stationary random sequence suppose had not been proved by previous reference, and that prediction did not present high precision, this paper proposes a method of fitting non-stationary random process probability distribution function.

Firstly, this paper proposes using weighted sum of Gauss items to fit zero-step approximate probability density. Secondly, for the beginning, this paper uses the method with few Gauss items under low precision. With the amount of points increasing, this paper uses more Gauss items under higher precision, and some Gauss items and some former points are deleted under precision condition. Thirdly, for particle swarm optimization with constraint problem, this paper proposed improved method, and the stop condition is under precision condition.

In experiment data analysis section, gear wear loss prediction is done by the method proposed by this paper. Compared with the method based on the stationary random sequence suppose by prediction relative error, the method proposed by this paper lowers the relative error whose absolute values are more than 5%, except when the current point sequence number is 2, and retains the relative error, whose absolute values are lower than 5%, still lower than 5%.

Finally, the method proposed by this paper based on non-stationary random sequence suppose is proved to be the better method in gear gravimetric wear loss prediction.

Due to the non‐stationarity of vibration signals resulting from either varying operating conditions or natural deterioration of machinery, both the frequency components and their magnitudes vary with time. However, little research has been done on the parameter estimation of time‐varying multivariate time series models based on adaptive filtering theory for condition‐based maintenance purposes. This paper proposes a state‐space model of non‐stationary multivariate vibration signals for the online estimation of the state of rotating machinery using a modified extended Kalman filtering algorithm and spectral analysis in the time‐frequency domain. Adaptability and spectral resolution capability of the model have been tested by using simulated vibration signal with abrupt changes and time‐varying spectral content. The implementation of this model to detect machinery deterioration under varying operating conditions for condition‐based maintenance purposes has been conducted by using real gearbox vibration monitoring signals. Experimental results demonstrate that the proposed model is able to quickly detect the actual state of the rotating machinery even under highly non‐stationary conditions with abrupt changes and yield accurate spectral information for an early warning of incipient fault in rotating machinery diagnosis. This is achieved through combination with a change detection statistic in bi‐spectral domain.