Pattern Recognition Algorithms Research Articles

Two newly developed indices evaluating facial skin aging.

Skin assessment methodologies have focused mainly on intuitive aging characteristics, including facial wrinkles and pigmented spots, and usually adopt pattern recognition algorithms. Recently, distinct methods of interpreting skin aging, such as the detection of facial landmarks and age prediction using machine learning techniques, have been conducted. We defined two indices that represent the severity of facial aging. The first index was the ratio of the bizygomatic distance and bigonial distance. The second index was the ratio of the degrees of the near mandible. The indices extracted from two-dimensional frontal face images were intended to show the deformation of the facial skin downward with aging progress. To validate whether these proposed indicators can represent facial aging, we conducted correlation tests with age and facial skin characteristics and performed association tests between the indices and facial skin characteristics, adjusted for age. The indices showed strong correlations with age (r=0.557 and 0.464, respectively) and facial skin characteristics. Although there were correlations between the indices and facial skin features, the associations between the indices and facial skin characteristics adjusted for age were weak or not significant. This suggests that the newly developed indices are appropriate for evaluating facial skin aging and distinct from typical measurements. We suggest two novel indices for evaluating facial aging based on frontal face images. The indices exhibited strong correlations with age and representative facial skin characteristics. The newly developed values can be differentiated indicators of facial aging compared with general skin features.

An Automatic Aerospace Tracking Data Analysis Method Based on Time Series Curve Shape

In the joint test of aerospace tracking, telemetering and command (TT&C) system, the consistency of measured and theoretical data is highly required. How to evaluate the quality of measured data intelligently and efficiently is a hot issue. To address this problem, an automatic data analysis method based on time series curve shape is proposed in this paper. By locating the part of abnormal shape on the curve, the data quality can be quickly evaluated using various pattern recognition algorithms. Finally, several simulation experiments are used to demonstrate the proposed method, which show that the curve shape can be accurately identified and the data quality can be given reasonably.

An AI-based Decision Support System for Predicting Mental Health Disorders.

Approximately one billion individuals suffer from mental health disorders, such as depression, bipolar disorder, schizophrenia, and anxiety. Mental health professionals use various assessment tools to detect and diagnose these disorders. However, these tools are complex, contain an excessive number of questions, and require a significant amount of time to administer, leading to low participation and completion rates. Additionally, the results obtained from these tools must be analyzed and interpreted manually by mental health professionals, which may yield inaccurate diagnoses. To this extent, this research utilizes advanced analytics and artificial intelligence to develop a decision support system (DSS) that can efficiently detect and diagnose various mental disorders. As part of the DSS development process, the Network Pattern Recognition (NEPAR) algorithm is first utilized to build the assessment tool and identify the questions that participants need to answer. Then, various machine learning models are trained using participants’ answers to these questions and other historical data as inputs to predict the existence and the type of their mental disorder. The results show that the proposed DSS can automatically diagnose mental disorders using only 28 questions without any human input, to an accuracy level of 89%. Furthermore, the proposed mental disorder diagnostic tool has significantly fewer questions than its counterparts; hence, it provides higher participation and completion rates. Therefore, mental health professionals can use this proposed DSS and its accompanying assessment tool for improved clinical decision-making and diagnostic accuracy.

Research on Degradation State Recognition of Axial Piston Pump under Variable Rotating Speed

Under the condition of variable rotating speed, it is difficult to extract the degradation characteristics of the axial piston pump, which also reduces the accuracy of degradation recognition. To address these problems, this paper proposes a degradation state recognition method for axial piston pumps by combining spline-kernelled chirplet transform (SCT), adaptive chirp mode pursuit (ACMP), and extreme gradient boosting (XGBoost). Firstly, SCT and ACMP are proposed to deal with the vibration signal instability and high noise of the axial piston pump under variable rotating speed. The instantaneous frequency (IF) of the axial piston pump can be extracted effectively by obtaining the accurate time-frequency distribution of signal components. Then, stable angular domain vibration signals are obtained by re-sampling, and multi-dimensional degradation characteristics are extracted from the angular domain and order spectrum. Finally, XGBoost is used to classify the selected characteristics to recognize the degradation state. In this paper, the vibration signals in four different degradation states are collected and analyzed through the wear test of the valve plate of the axial piston pump. Compared with different pattern recognition algorithms, it is verified that this method can ensure high recognition accuracy.

One metal-ion-regulated AgTNPs etching sensor array for visual discrimination of multiple organic acids

The detection and discrimination of organic acids (OAs) is of great importance in the early diagnosis of specific diseases. In this study, we established an effective visual sensor array for the identification of OA. This is the first time, to our best knowledge, that metal ions were used to regulate the etching of silver triangular nanoprisms (AgTNPs) in an OA discrimination sensor array. The sensor array was based on the oxidation etching of AgTNPs by three metal ions (Mn2+, Pb2+, and Cr3+) and accelerated etching of AgTNPs by OA. The introduction of metal ions alone led to a slight wavelength shift of the AgTNPs colloid solution, signifying the incomplete etching of the AgTNPs. Nevertheless, when metal ions and OA were introduced simultaneously to the solution, a significant blueshift of the localized surface plasmon resonance peak was detected, and a color change of the AgTNPs was observed, which were the consequences of morphological transitions of the AgTNPs. The addition of different OA accelerated AgTNPs etching in varying degrees, generating diverse colorimetric response patterns (i.e., RGB variations) as "fingerprints" associated with each specific organic acid. Pattern recognition algorithms and neural network simulation were employed to further data analysis, indicating the outstanding discrimination capability of the provided array for eight OA at the 33 µM level. Moreover, excellent results of selective experiments as well as real samples tests demonstrate that our proposed method possesses great potential for practical applications.

A supervised discriminant data representation: application to pattern classification

The performance of machine learning and pattern recognition algorithms generally depends on data representation. That is why, much of the current effort in performing machine learning algorithms goes into the design of preprocessing frameworks and data transformations able to support effective machine learning. The method proposed in this work consists of a hybrid linear feature extraction scheme to be used in supervised multi-class classification problems. Inspired by two recent linear discriminant methods: robust sparse linear discriminant analysis (RSLDA) and inter-class sparsity-based discriminative least square regression (ICS_DLSR), we propose a unifying criterion that is able to retain the advantages of these two powerful methods. The resulting transformation relies on sparsity-promoting techniques both to select the features that most accurately represent the data and to preserve the row-sparsity consistency property of samples from the same class. The linear transformation and the orthogonal matrix are estimated using an iterative alternating minimization scheme based on steepest descent gradient method and different initialization schemes. The proposed framework is generic in the sense that it allows the combination and tuning of other linear discriminant embedding methods. According to the experiments conducted on several datasets including faces, objects, and digits, the proposed method was able to outperform competing methods in most cases.

Near-infrared spectroscopy combined with pattern recognition algorithms to quickly classify raisins

With the development of commodity economy, the emergence of fake and shoddy raisin has seriously harmed the interests of consumers and enterprises. To deal with this problem, a classification method combining near-infrared spectroscopy and pattern recognition algorithms were proposed for adulterated raisins. In this study, the experiment was performed by three kinds of raisins in Xinjiang (Hongxiangfei, Manaiti, Munage). After collecting and normalizing the spectral data, we compared the spectra of three kinds of raisins. Next the principal component analysis (PCA) was preformed to compress the dimension of the spectral data, and then classification models including support vector machine (SVM), multiscale fusion convolutional neural network (MCNN) and improved AlexNet were established to identify raisins. The accuracy of SVM, MCNN, and improved AlexNet is 100%, 92.83%, and 97.78% respectively. This study proves that near-infrared spectroscopy combined with pattern recognition is feasible for the raisin inspection.

Enhancing the printing accuracy of melt electrowritten fibers deposited on aluminum foils

Melt electrowriting (MEW) is an emerging extrusion-based additive manufacturing technology to process high-fidelity 3D scaffold. To date, the aluminum foil is a commonly-applied substrate to collect the melt electrowritten fibers. However, inaccurate fiber deposition cannot be efficiently avoided when printing on the aluminum foils. In this study, the effects of placing angle of the aluminum foil on the fiber deposition accuracy was investigated. In addition, to eliminate the human error, a planar pattern recognition algorithm was developed to evaluate the fiber deposition accuracy. It was found that the interactions between the fibers and the texture of the aluminum foils would influence the fiber deposition accuracy. The aluminum foil with 45° relative placement angle exhibited the smallest fiber deposition deviation ratio (less than 1.16%), which indicating that reducing the parallelism between the aluminum foil texture and the printed fiber path can significantly improve the printing resolution of MEW scaffold.

Detection of the Driver’s Mental Workload Level in Smart and Autonomous Systems Using Physiological Signals

With the continuous advancement of automation technology, autonomous driving assistance systems are gradually sharing the tasks during driving, but the driver still assumes the main driving tasks. In addition to driving activities, the advent of numerous new functions will have an indirect impact on the driver’s mental effort. However, determining the driver’s mental effort remains a difficult issue. In this paper, a method is proposed to assess the mental workload of drivers, combining real driver’s physiological data with the speed of his/her vehicle. The correlation coefficient and significance level are obtained by analyzing the correlation between physiological data and road types. The relevant data is then preprocessed to determine the characteristic index, with the mental workload as the input index. The driver’s mental workload is classified and the mental workload prediction model is constructed on the basis of the combination of the Fuzzy Pattern Recognition Algorithm and Genetic Algorithm. At the same time, the suggested approach is compared to the J48 Classification Algorithm and the Simulated Annealing Optimization Algorithm. The results demonstrate that the proposed method in this paper’s effectiveness for identifying the driver’s mental workload level is evidently better than other algorithms, which provides new theoretical support for assessing the L3+ driver’s mental workload level under the background of the safety of the intended functionality when they take over the control of the drive.

Deep Learning (CNN) and Transfer Learning: A Review

Deep Learning is a machine learning area that has recently been used in a variety of industries. Unsupervised, semi-supervised, and supervised-learning are only a few of the strategies that have been developed to accommodate different types of learning. A number of experiments showed that deep learning systems fared better than traditional ones when it came to image processing, computer vision, and pattern recognition. Several real-world applications and hierarchical systems have utilised transfer learning and deep learning algorithms for pattern recognition and classification tasks. Real-world machine learning settings, on the other hand, often do not support this assumption since training data can be difficult or expensive to get, and there is a constant need to generate high-performance beginners who can work with data from a variety of sources. The objective of this paper is using deep learning to uncover higher-level representational features, to clearly explain transfer learning, to provide current solutions and evaluate applications in diverse areas of transfer learning as well as deep learning.

Development and application of a supervised pattern recognition algorithm for identification of fuel-specific emissions profiles

Abstract. Wildfires have increased in frequency and intensity in the western United States (US) over the past decades, with negative consequences for air quality. Wildfires emit large quantities of particles and gases that serve as air pollutants and their precursors, and can lead to severe air quality conditions over large spatial and long temporal scales. Therefore, characterization of the chemical constituents in smoke as a function of combustion conditions, fuel type and fuel component is an important step towards improving the prediction of air quality effects from fires and evaluating mitigation strategies. Building on the comprehensive characterization of gaseous non-methane organic compounds (NMOCs) identified in laboratory and field studies, a supervised pattern recognition algorithm was developed that successfully identified unique chemical speciation profiles among similar fuel types common in western coniferous forests. The algorithm was developed using laboratory data from single fuel species and tested on simplified synthetic fuel mixtures. The fuel types in the synthetic mixtures were differentiated, but as the relative mixing proportions became more similar, the differentiation became poorer. Using the results from the pattern recognition algorithm, a classification model based on linear discriminant analysis was trained to differentiate smoke samples based on the contribution(s) of dominant fuel type(s). The classification model was applied to field data and, despite the complexity of the contributing fuels and the presence of fuels “unknown” to the classifier, the dominant sources/fuel types were identified. The pattern recognition and classification algorithms are a promising approach for identifying the types of fuels contributing to smoke samples and facilitating the selection of appropriate chemical speciation profiles for predictive air quality modeling using a highly reduced suite of measured NMOCs. The utility and performance of the pattern recognition and classification algorithms can be improved by expanding the training and test sets to include data from a broader range of single and mixed fuel types.

SVM Classification Method of Waxy Corn Seeds with Different Vitality Levels Based on Hyperspectral Imaging

The vitality of corn seeds is a significant indicator for assessing the quality and yield of crops. In recent years, numerous information technologies have been adopted to analyze the seed vitality and provide support for efficient equipment. However, there are still some shortcomings in these technologies, which decrease the accuracy of identifying the seed vitality for various practical applications. In this paper, a synthesized classification method for seed vitality was proposed based on multisensor hyperspectral imaging. Firstly, hyperspectral images in the range of 370-1042 nm were collected for waxy corn seeds, which were subjected to aging processing with four periods of time (0, 3, 6, and 9 d). Besides, some preprocessing techniques including standard normal variate, multiplicative scatter correction, Savitzky-Golay smoothing, and first-order and second-order derivatives were employed to suppress noise interference in raw spectra. In addition, principal component analysis (PCA), 2nd derivatization, and successive projection algorithm (SPA) were adopted to select feature wavelengths. Moreover, SVM classification models based on full spectra and feature wavelengths were established. The results showed that, based on feature wavelengths selected by SPA, the SVM model preprocessed by multiplicative scatter correction (MSC) had the optimal performance. The training accuracy and testing accuracy of this model were 100% and 97.9167%, respectively. RMSE was 0.018 and R 2 was 0.875. Therefore, it can be demonstrated that the pattern recognition algorithm could achieve a high accuracy in classifying accelerated aging seeds. This algorithm provides a new method for machine learning (ML) in nondestructive detection of crops.

Local tsunamigenic sources in Greece, identified by pattern recognition

This study is an attempt to determine potential tsunamigenic morphostructural nodes in Greece using pattern recognition algorithms. The earthquakes that have produced local tsunamis in the region were confined to morphostructural nodes whose locations were found by morphostructural zoning. The recognition problem consisted in separating all nodes in the region into the tsunamigenic class and the non-tsunamigenic class based mainly on the geomorphologic parameters of the nodes. The data on tsunamigenic earthquakes in Greece for training the Cora-3 algorithm were taken from the GHTD global historical catalog of tsunamigenic events (http://www.ngdc.noaa.gov/hazard/tsu_db.shtml). The recognition procedure resulted in determining 27 tsunamigenic nodes, with most of these being situated in the southern tip of the Peloponnese Peninsula, as well as in the gulfs of Corinth and Patras. Three tsunamigenic nodes were identified in the area of the Malian Gulf on the Aegean coast of Greece. According to the relevant literature, most local tsunamis in Greece were initiated by submarine slides and slumps due to earthquakes. According to the characteristic geomorphologic features derived in this study, the tsunamigenic nodes are situated in settings of contrasting relief characterized by steep slopes. This favors submarine landslides when subjected to earthquake excitation. The results reported in this paper form a basis for developing a methodology to be used in long-term tsunami hazard assessment, supplying information on local potential tsunamigenic sources required for tsunami regionalization of coastal areas in Greece.

Neural Network Robustness Analysis Using Sensor Simulations for a Graphene-Based Semiconductor Gas Sensor

Despite their advantages regarding production costs and flexibility, chemiresistive gas sensors often show drawbacks in reproducibility, signal drift and ageing. As pattern recognition algorithms, such as neural networks, are operating on top of raw sensor signals, assessing the impact of these technological drawbacks on the prediction performance is essential for ensuring a suitable measuring accuracy. In this work, we propose a characterization scheme to analyze the robustness of different machine learning models for a chemiresistive gas sensor based on a sensor simulation model. Our investigations are structured into four separate studies: in three studies, the impact of different sensor instabilities on the concentration prediction performance of the algorithms is investigated, including sensor-to-sensor variations, sensor drift and sensor ageing. In a further study, the explainability of the machine learning models is analyzed by applying a state-of-the-art feature ranking method called SHAP. Our results show the feasibility of model-based algorithm testing and substantiate the need for the thorough characterization of chemiresistive sensor algorithms before sensor deployment in order to ensure robust measurement performance.

A Novel Approach to Using Seabed Geomorphology as a Predictor of Habitat Use in Highly Mobile Marine Predators: Implications for Ecology and Conservation

Understanding how marine predators find patchily distributed prey resources in a dynamic environment is key to identifying important ecological areas for ecosystem-level conservation management. However, the mechanisms underpinning important foraging areas often result from complex interactions between static and dynamic covariates (e.g. topography and currents). Modelling habitat associations with hydrodynamic processes is rarely useful when attempting to identify and characterise foraging areas across an individual’s foraging range. Investigating the influence of static habitat features on predator behaviour can provide a more tractable baseline understanding of habitat associations, upon which additional complexity can be added. Seabed gradient covariates (e.g. slope and aspect) are often used, yet such metrics are computed at singular user-defined resolutions, and provide limited ecological insight when used in isolation. Instead, categorising the seabed into geomorphological features may provide better characterisation of seabed structure. Here we explore the utility of a pattern recognition algorithm to delineate whole geomorphological features (“geomorphons”) on the seabed (e.g. valleys, ridges, footslopes) from bathymetry data, and examine the influence of geomorphology on marine predator habitat use. We demonstrate the potential application of this approach in a case study, examining the influence of geomorphons on the at-sea behaviour of a highly mobile predator inhabiting shelf seas: the grey seal (Halichoerus grypus). We analyse GPS tracking data from three seals tagged in the southern North Sea, an area with heterogeneous geomorphology. We use hidden Markov models (HMMs) to infer foraging and travelling behaviour and model the effect of different feature types on the probability of switching between states. All three seals showed an increased probability of transitioning from travelling to foraging when encountering slopes, footslopes and hollows, and foraging activity was concentrated at slopes on the fringes of the Dogger Bank. We hypothesise that such features may host prey aggregations, and/or lead to increased prey capture success. The results suggest the importance of such areas for grey seals in the southern North Sea, a region undergoing rapid and widespread anthropogenic habitat change. This method could be incorporated into future species distribution models to improve estimates of predator distribution, informing conservation management and marine spatial planning.

ANN based approach for selective detection of breath acetone by using hybrid GO-FET sensor array

This research used hybrid graphene oxide (GO) field effect transistors (FETs) based sensor array to design an electronic nose (e-nose) for identifying exhaled breath acetone to diagnose diabetes mellitus through noninvasive route. Six back gated FET sensors were fabricated with hybrid channel of GO, WO3 and noble metals (Au, Pd and Pt) nanoparticles. The experiment was carried out by using four distinct forms of synthetic breath, each with a different level of interference. Linear discriminant analysis (LDA) and artificial neural networks (ANN) were utilized to classify and analyze the sensor response vector. In contrast, partial least square (PLS) and multiple linear regression (MLR) were used to evaluate the exact acetone concentration in synthetic breath. First, LDA was used to lower the dimensionality of the response vector, which was then provided as an input to the ANN model. ANN was performed with ten perceptrons model in the hidden layer and highest accuracy of 99.1% was achieved. Additionally, by using the loading plot of PLS, three sensors (Pt/WO3/GO, Pd/WO3/GO, and WO3/GO) had the ample use to predict the concentration of breath acetone. Moreover, the MLR approach with correlation coefficient (R2) of 0.9572 and root mean square error (RMSE) of 5.63% were used for obtaining the exact concentration of acetone. Consequently, e-nose with matrix of hybrid GO-FET sensors and pattern recognition algorithms (LDA, ANN, PLS and MLR) exhibited considerable ability in selective detection of acetone in synthetic breath.

Ensemble machine learning approach for electronic nose signal processing

Electronic nose (e-nose) systems have been reported to be used in many areas as rapid, low-cost, and non-invasive instruments. Especially in meat production and processing, e-nose system is a powerful tool to process volatile compounds as a unique ‘fingerprint’. The ability of the pattern recognition algorithm to analyze e-nose signals is the key to the success of the e-nose system in many applications. On the other hand, ensemble methods have been reported for favorable performances in various data sets. This research proposes an ensemble learning approach for e-nose signal processing, especially in beef quality assessment. Ensemble methods are not only used for learning algorithms but also sensor array optimization. For sensor array optimization, three filter-based feature selection algorithms (FSAs) are used to build ensemble FSA such as reliefF, chi-square, and gini index. Ensemble FSA is developed to deal with different or unstable outputs of a single FSA on homogeneous e-nose data sets in beef quality monitoring. Moreover, ensemble learning algorithms are employed to deal with multi-class classification and regression tasks. Random forest and Adaboost are used that represent bagging and boosting algorithms, respectively. The results are also compared with support vector machine and decision tree as single learners. According to the experimental results, our ensemble approach has good performance and generalization in e-nose signal processing. Optimized sensor combination based on filter-based FSA shows stable results both in classification and regression tasks. Furthermore, Adaboost as a boosting algorithm produces the best prediction even though using a smaller number of sensors.

A training pattern recognition algorithm based on weight clustering for improving cooling load prediction accuracy of HVAC system

The cooling load-based optimal control is an advanced technology for the efficient operation of heating, ventilation, and air conditioning (HVAC). Thus, the prediction reliability of cooling load plays a key role in HVAC's optimal control. Current publications primarily focused on the structure optimization of prediction models, while less on the clustering-based cooling load prediction. However, the data quality determines the upper limit of the model's prediction performance. Thus, a training pattern recognition algorithm based on weight clustering is proposed for improving cooling load prediction accuracy. Compared with the existing clustering-based prediction methods, the main innovations of the proposed method are: (i) considering the input variables' weights on cooling load in the clustering process; and (ii) investigating the matching between the various prediction models and the K-means clustering algorithm. The case studies showed that the proposed method achieves a significant improvement in the prediction performance, such as MAPEs of the MLR, MNR, and ANN decrease by 34.67%, 35.56%, and 14.53% on average, respectively. Compared with the non-weights clustering method, the introduction of the weights can further improve the above models' prediction accuracy, such as their MAPEs decrease by 6.30%, 7.59%, and 3.07% on average, respectively. These results also demonstrated that the clustering-based prediction method is more suitable for the regression models (e.g., MLR and MNR) with low complexity compared to the ANN. When the clustering number is about 4, the models' prediction performances were more robust. Applying the proposed method to the time-series models (i.e., AR, ARX, and ANN) resulted in their MAPEs as low as 1.79%, 1.78%, and 2.06%, respectively. the proposed method can provide a new idea for improving the accuracy of cooling load prediction.

Application of individualized differential expression analysis in human cancer proteome.

Liquid chromatography-mass spectrometry-based quantitative proteomics can measure the expression of thousands of proteins from biological samples and has been increasingly applied in cancer research. Identifying differentially expressed proteins (DEPs) between tumors and normal controls is commonly used to investigate carcinogenesis mechanisms. While differential expression analysis (DEA) at an individual level is desired to identify patient-specific molecular defects for better patient stratification, most statistical DEP analysis methods only identify deregulated proteins at the population level. To date, robust individualized DEA algorithms have been proposed for ribonucleic acid data, but their performance on proteomics data is underexplored. Herein, we performed a systematic evaluation on five individualized DEA algorithms for proteins on cancer proteomic datasets from seven cancer types. Results show that the within-sample relative expression orderings (REOs) of protein pairs in normal tissues were highly stable, providing the basis for individualized DEA for proteins using REOs. Moreover, individualized DEA algorithms achieve higher precision in detecting sample-specific deregulated proteins than population-level methods. To facilitate the utilization of individualized DEA algorithms in proteomics for prognostic biomarker discovery and personalized medicine, we provide Individualized DEP Analysis IDEPAXMBD (XMBD: Xiamen Big Data, a biomedical open software initiative in the National Institute for Data Science in Health and Medicine, Xiamen University, China.) (https://github.com/xmuyulab/IDEPA-XMBD), which is a user-friendly and open-source Python toolkit that integrates individualized DEA algorithms for DEP-associated deregulation pattern recognition.

Application of a Nonparametric Pattern Recognition Algorithm to the Problem of Testing the Hypothesis of the Independence of Variables of Multi-Valued Functions

Application of a Nonparametric Pattern Recognition Algorithm to the Problem of Testing the Hypothesis of the Independence of Variables of Multi-Valued Functions