Feature identification in complex fluid flows by convolutional neural networks

Recent advancements have established machine learning's utility in predicting nonlinear fluid dynamics,with predictive accuracy being a central motivation for employing neural networks.However,the pattern recognition central to the networks function is equally valuable for enhancing our dynamical insight into the complex fluid dynamics.In this paper,a single-layer convolutional neural network(CNN)was trained to recognize three qualitatively different subsonic buffet flows(periodic,quasi-periodic and chaotic)over a high-incidence airfoil,and a near-perfect accuracy was obtained with only a small training dataset.The convolutional kernels and corresponding feature maps,developed by the model with no temporal information provided,identified large-scale coherent structures in agreement with those known to be associated with buffet flows.Sensitivity to hyperparameters including network architecture and convolutional kernel size was also explored.The coherent structures identified by these models enhance our dynamical understanding of subsonic buffet over high-incidence airfoils over a wide range of Reynolds numbers.

Stability-mutation feature identification of Web search keywords based on keyword concentration change ratio

Purpose: The aim of this paper is to discuss how the keyword concentration change ratio(KCCR) is used while identifying the stability-mutation feature of Web search keywords during information analyses and predictions.Design/methodology/approach: By introducing the stability-mutation feature of keywords and its significance, the paper describes the function of the KCCR in identifying keyword stability-mutation features. By using Ginsberg's influenza keywords, the paper shows how the KCCR can be used to identify the keyword stability-mutation feature effectively.Findings: Keyword concentration ratio has close positive correlation with the change rate of research objects retrieved by users, so from the characteristic of the 'stability-mutation' of keywords, we can understand the relationship between these keywords and certain information. In general, keywords representing for mutation fit for the objects changing in short-term, while those representing for stability are suitable for long-term changing objects. Research limitations: It is difficult to acquire the frequency of keywords, so indexes or parameters which are closely related to the true search volume are chosen for this study.Practical implications: The stability-mutation feature identification of Web search keywords can be applied to predict and analyze the information of unknown public events through observing trends of keyword concentration ratio.Originality/value: The stability-mutation feature of Web search could be quantitatively described by the keyword concentration change ratio(KCCR). Through KCCR, the authors took advantage of Ginsberg's influenza epidemic data accordingly and demonstrated how accurate and effective the method proposed in this paper was while it was used in information analyses and predictions.

An online identity authentication method for blood smear

Blood smear test is the basic method of blood cytology and is also a standard medical test that can help diagnose various conditions and diseases.Morphological examination is the gold stan-dard to determine pathological changes in blood cell morphology.In the biology and medicine automation trend,blood smears'automated management and analysis is very necessary.An online blood smear automatic microscopic image detection system has been constructed.It includes an online blood smear automatic producing part and a blood smear automatic micro-scopic image detection part.Online identity authentication is at the core of the system.The identifiers printed online always present dot matrix digit code(DMDC)whose stroke is not continuous.Considering the particularities of DMDC and the complexities of online application environment,an online identity authentication method for blood smear with heterological theory is proposed.By synthesizing the certain regional features according to the heterological theory,high identification accuracy and high speed have been guaranteed with few features required.In the experiment,the suficient correct matches bet ween the tube barcode and the identification result verified its feasibility and validity.

The admittance feature representation and impact sound feature extraction in the material identification of ribbed plates

The admittance features representing the physical attributes are used as the in termediates to extract the materialattributesrelated impact sound features of ribbed plates. Firstly, the admittance feature representations of metal ribbed plates attributes are obtained and the relationship between the admittance features and the impact sound features are established via correlation analysis method. Then, materialattributesrelated impact sound features are obtained indirectly. Finally, the performances of different sound features for the material recognition of ribbedmetal plates are verified through the Support Vector Machine classifier. The results indicate that the obtained four sets of features can effectively identify the materials of the metal ribbed plates, while the accuracy of a single feature depends on the separable degree of the corresponding material attribute. And the features extracted based on admittance functions have higher average accuracy than that of timbre features. Therefore, the proposed sound feature extraction method based on admittance features is valid, and the extracted sound features can effectively reflect the physical attributes.

A hybrid data-mining framework for train rescheduling strategy pattern discovery

This study presents a hybrid data-mining framework based on feature selection algorithms and clustering methods to perform the pattern discovery of high-speed railway train rescheduling strategies(RSs).The proposed model is composed of two states.In the first state,decision tree,random forest,gradient boosting decision tree(GBDT)and extreme gradient boosting(XGBoost)models are used to investigate the importance of features.The features that have a high influence on RSs are first selected.In the second state,a K-means clustering method is used to uncover the interdependences between RSs and the influencing features,based on the results in the first state.The proposed method can determine the quantitative relationships between RSs and influencing factors.The results clearly show the influences of the factors on RSs,the possibilities of different train operation RSs under different situations,as well as some key time periods and key trains that the controllers should pay more attention to.The research in this paper can help train traffic controllers better understand the train operation patterns and provides direction for optimizing rail traffic RSs.

Peripheral Nonlinear Time Spectrum Features Algorithm for Large Vocabulary Mandarin Automatic Speech Recognition

This work describes an improved feature extractor algorithm to extract the peripheral features of point x(ti,fj) using a nonlinear algorithm to compute the nonlinear time spectrum (NL-TS) pattern. The algo- rithm observes n×n neighborhoods of the point in all directions, and then incorporates the peripheral fea- tures using the Mel frequency cepstrum components (MFCCs)-based feature extractor of the Tsinghua elec- tronic engineering speech processing (THEESP) for Mandarin automatic speech recognition (MASR) sys- tem as replacements of the dynamic features with different feature combinations. In this algorithm, the or- thogonal bases are extracted directly from the speech data using discrite cosime transformation (DCT) with 3×3 blocks on an NL-TS pattern as the peripheral features. The new primal bases are then selected and simplified in the form of the ?dp- operator in the time direction and the ?dp- operator in the frequency di- t f rection. The algorithm has 23.29% improvements of the relative error rate in comparison with the standard MFCC feature-set and the dynamic features in tests using THEESP with the duration distribution-based hid- den Markov model (DDBHMM) based on MASR system.