Support Vector Machine Research Articles

An explainable deep learning approach for stock market trend prediction

Given the intricate nature of stock forecasting as well as the inherent risks and uncertainties, analysis of market trends is necessary to capitalize on optimal investment opportunities for profit maximization and timely disinvestment for loss minimization. In this work, we propose a deep learning model for predicting five distinct stock market trends: upward, downward, double top, rounded bottom, and rounded top. The proposed model surpasses common benchmarks, including support vector machine, random forest, and logistic regression, achieving an average accuracy of 94.9%, compared to 85.7% for random forest, 60.07% for support vector machine, and 52.45% for logistic regression. Furthermore the proposed model excels in F1-score, with a 94.85% performance, compared to 77.95% for random forest, 21.02% for support vector machine and 46.23% for logistic regression, across four real world diverse datasets. Additionally, we employ explainable AI (XAI) techniques, SHAP and LIME, to enhance interpretability, enabling stakeholders to understand the key factors driving predictions. The SHAP analysis reveals the top 10 most important/influential features, enabling feature reduction while maintaining performance. Interestingly, while accuracy slightly decreases with top 10 features, precision, recall, and F1-score improve, suggesting a trade-off between comprehensiveness and performance. These results demonstrate the potential for practical application in financial decision-making, providing a balance between interpretability and predictive power that can support investors in risk management and strategic planning.

Machine Learning-Based Cow Milk Quality Classification using Recursive Feature Elimination Cross-Validation

<p>Milk quality is of paramount importance as it directly impacts consumer health and well-being. High-quality milk is rich in essential nutrients such as calcium, protein, and vitamins, contributing to overall nutrition. Moreover, ensuring milk quality is vital for preventing the transmission of diseases and contaminants through dairy products. Therefore, research in this field is essential to guaranteeing the safety and nutritional value of milk consumed by individuals of all ages. In this paper, the design of machine learning-based grade measuring devices with recursive feature elimination with cross-validation (RFECV) is carried out as a guide in the design of a milk grade detection system. The milk is rated as low, medium, or high based on these criteria. The sensors will gather this information from the milk with the aid of the microcontroller. The algorithms utilized in this study and the results obtained from K-Nearest Neighbors (KNN) combined with the RFECV algorithm have a higher accuracy value: 17,20% better than the support vector machine (SVM) model, 25.37% better than the single K-Nearest Neighbors (KNN), and 26.37% better than the random forest (RF) model trained without RFECV. Using seven input features (pH, temperature, taste, odor, fat, turbidity, and color), the proposed model produces 96.27% accuracy.</p>

Bayesian additive regression trees for predicting childhood asthma in the CHILD cohort study

BackgroundAsthma is a heterogeneous disease that affects millions of children and adults. There is a lack of objective gold standard diagnosis that spans the ages; instead, diagnoses are made by clinician assessment based on a cluster of signs, symptoms and objective tests dependent on age. Yet, there is a clear morbidity associated with chronic asthma symptoms. Machine learning has become a popular tool to improve asthma diagnosis and classification. There is a paucity of literature on the use of Bayesian machine learning algorithms to predict asthma diagnosis in children. This paper develops a prediction model using the Bayesian additive regression trees (BART) and compares its performance to various machine learning algorithms in predicting the diagnosis of childhood asthma.MethodsClinically relevant variables collected at or before 3 years of age from 2794 participants in the CHILD Cohort Study were used to predict physician-diagnosed asthma at age 5. BART and six other commonly used machine learning algorithms, namely adaptive boosting, logistic regression, decision tree, neural network, random forest, and support vector machine were trained. Measures of performance including sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve were calculated. The confidence intervals were calculated using Bootstrapping samples. Important predictors and interaction effects associated with asthma were also identified using BART.ResultsBART, logistic regression and random forest showed the highest area under the ROC curve compared to other machine learning algorithms. Based on BART, recurrent wheeze, respiratory infection and food sensitization at 3 years of age were the most important predictors. The three most important interaction effects were found to be interaction terms of respiratory infection at 3 years and recurrent wheezing at 3 years, maternal asthma and paternal asthma, and maternal wheezing and inhalant sensitization of child at 3 years.ConclusionsBART demonstrated promising prediction performance when compared to other machine learning algorithms. Future research could validate the BART in an external cohort to evaluate its reliability and generalizability.

Combination of plasma-based lipidomics and machine learning provides a useful diagnostic tool for ovarian cancer

Ovarian cancer (OC), the second leading cause of death among gynecologic cancers, is often diagnosed at an advanced stage due to its asymptomatic nature at early stages. This study aimed to explore the diagnostic potential of plasma-based lipidomics combined with machine learning (ML) in OC. Non-targeted lipidomics analysis was conducted on plasma samples from participants with epithelial ovarian cancer (EOC), benign ovarian tumor (BOT), and healthy control (HC). The samples were randomly divided into a train set and a test set. Differential lipids between groups were selected using two-tailed Student’s t-test and partial least squares discriminant analysis (PLS-DA). Both single lipid-based receiver operating characteristic (ROC) model, and multiple lipid-based ML model, were constructed to investigate the diagnostic value of the differential lipids. The results showed several lipids with significant diagnostic potential. ST 27:2;O achieved the highest prediction accuracy of 0.92 in distinguishing EOC from HC. DG 42:2 had the highest prediction accuracy of 0.96 in diagnosing BOT from HC. Cer d18:1/18:0 had the highest prediction accuracy of 0.65 in differentiating EOC from BOT. Furthermore, multiple lipid-based ML models illustrated better diagnostic performance. K-nearest neighbors (k-NN), partial least squares (PLS), and random forest (RF) models achieved the highest prediction accuracy of 0.96 in discriminating EOC from HC. The support vector machine (SVM) model reached the highest prediction accuracy both in distinguishing BOT from HC, and in differentiating EOC from BOT, with accuracies of 1.00 and 0.74, respectively. In conclusion, this study revealed that the combination of plasma-based lipidomics and ML algorithms is an effective method for diagnosing OC.

Prediction of fracture toughness of concrete using the machine learning approach

In the process of structural design, it is useful to estimate the fracture toughness of concrete samples. This research showcases the effectiveness of utilizing machine learning methods to determine the fracture toughness of concrete. Taking into account variables such as mix design, machine learning techniques can accurately predict the mode I fracture toughness of concrete. Dimensionless stress intensity factor of concrete prediction using twelve different machine learning techniques namely, Linear regression (LR), Extreme Gradient-Boosting (XGboost), K-Nearest Neighbors (KNN), Random Forest (RF), Category Boosting (CB), Decision Tree (DT), Extra Trees (ET), Light Gradient-Boosting (LightGB), Adaptive boosting (AdaBoost), Bagging (BA), Gaussian Process (GP), Artificial Neural Network (ANN) and Support Vector Machine (SVM). The result of utilizing the training adaptive moment estimation algorithm has been developed to create an outstanding machine learning-based system. After carefully analyzing comparisons between the predictions produced by different models and experimental findings, it has been discovered that the models demonstrate an impressive accuracy rate of about 90 percent when it comes to forecasting concrete fracture toughness. The research findings emphasize that the ANN model exhibited superior accuracy in its predictions (R2 value of 0.90, RMSE of 0.1517, and MAE of 0.1238). Upon conducting a thorough examination of the ANN method’s sensitivity, the cement parameter holds utmost significance in accurately estimating concrete’s fracture toughness using the available dataset. So, the ANN model can be used as a valuable method to provide practical assistance in predicting the fracture toughness of concrete. When evaluating the effective parameters, the cement and metakaolin dosage and the notch height to specimen height ratio have the greatest effect on the fracture resistance, while the coarse aggregate content is minimal. This result is consistent with the experimental data.

Predictive Modelling of Mineral Prospectivity Using Satellite Remote Sensing and Machine Learning Algorithms

In today’s world of falling returns on fixed exploration budgets, complex targets, and ever-increasing volumes of multi-parameter datasets, the effective management and integration of existing data are essential to any mineral exploration operation. Machine learning (ML) algorithms like Convolutional Neural Networks (CNN), Random Forest (RF), and Support Vector Machine (SVM) are powerful data-driven methods that are not implemented very often with remote sensing-derived hydrothermal alternation information and limited field datasets for mapping mineral prospectivity. The application of machine learning algorithms with satellite remote sensing data and limited field data, they have not been compared and evaluated together thoroughly in this field. A data science approach was applied to create nine predictor maps, incorporating limited field data and satellite remote sensing information. A confusion matrix, statistical measures, and a Receiver Operating Characteristic (ROC) curve were used to evaluate the prediction models efficacy on both the training and test datasets. The results suggested that the RF model exhibited the highest predictive accuracy, consistency and interpretability among the three ML models evaluated in this study. RF model also achieved the highest predictive efficiency in capturing known copper (Cu) deposits within a small prospective area. In comparison to the SVM and CNN models, the RF model outperformed them in terms of predictive accuracy and interpretability. These results imply that the RF model is the most suitable for Cu potential mapping in the Pakistan’s North Waziristan region. Consequently, all the models including the RF model were used to generate a prospectivity map, which contained low to very-high potential zones, to support further exploration in the region. The newly discovered deposit inside the predicted prospective areas demonstrates the robustness and efficacy of the prospectivity modelling approach as proposed in this research for generating exploration targets.

Research on an aeroengine bearing fault diagnosis method based on WMSST-CNN-BKA-LSSVM

Abstract To effectively extract the fault feature from small sample datasets and improve the accuracy of the model, an aeroengine bearing fault diagnosis method is proposed in this research based on wavelet multi-synchrosqueezed transform-convolutional neural network-black-winged kite algorithm-least squares support vector machine (WMSST-CNN-BKA-LSSVM). WMSST was used to perform modal decomposition on the collected fault data to obtain the time-frequency images. The time-frequency images were input into CNN for fault data feature extraction. The results of the CNN fully connected layer were used as input to the LSSVM, and the BKA was used to optimize the key parameters of the LSSVM to classify aeroengine bearing faults. The method was verified by aeroengine bearing test data from the Harbin Institute of Technology. The research showed that the accuracy is 99.86% when the faulty test samples are 70%, and the accuracy is 93.90% when the faulty samples are 30%. The accuracy showed the applicability of this method in small sample fault diagnosis. By comparing the method with a one-dimensional convolutional neural network (1D-CNN) and the convolutional neural network - long short-term memory (CNN-LSTM), it is found that the accuracy has been dramatically improved.

Utilizing Saliva Metabolomics for Diagnosing Gastric Cancer and Exploring the Changes in Saliva Metabolites After Surgery.

Gastric cancer (GC) is a disease with high prevalence and mortality, but we lack convenient and accurate methods to screen for this disease. Thus, we aimed to search for some salivary biomarkers and explore changes in metabolites in patients' saliva after radical gastrectomy. A total of 152 subjects were divided into three groups (healthy group, GC group, and one-week postoperative group). After simple processing, saliva samples were analyzed by liquid chromatography-mass spectrometry. First, we used total ion chromatography and principal component analysis to determine the metabolite profiles. Next, t-test, partial least squares discriminant analysis, support vector machine, and receiver operating characteristics curve analysis were performed to identify biomarkers. Then, Fisher discriminant analysis and hierarchical clustering analysis were performed to determine the discriminating ability of biomarkers. Finally, we established a generalized linear model to predict GC based on biomarkers, and used bootstrapping for internal validation. Between the healthy and GC groups, we identified four biomarkers: lactic acid, kynurenic acid, 3-hydroxystachydrine, and S-(1,2,2-trichlorovinyl)-L-cysteine. We used stepwise regression to select five metabolites and develop a model with areas under the curve equal to 0.973 in the training dataset and 0.924 in the validation dataset. Between the GC and one-week postoperative groups, we found two differential metabolites: 19-hydroxyprostaglandin E2 and DG (14:0/0:0/18:2n6). Differential metabolites were observed among the three groups. GC could be initially diagnosed on the basis of detection of these biomarkers. Moreover, changes in salivary metabolites in postoperative patients could provide important insights for basic studies.

Abnormal intrinsic brain functional network dynamics in patients with retinal detachment based on graph theory and machine learning

Backgroundand purpose: The investigation of functional plasticity and remodeling of the brain in patients with retinal detachment (RD) has gained increasing attention and validation. However, the precise alterations in the topological configuration of dynamic functional networks are still not fully understood. This study aimed to investigate the topological structure of dynamic brain functional networks in RD patients. MethodsWe recruited 32 patients with RD and 33 healthy controls (HCs) to participate in resting-state fMRI. Employing the sliding time window analysis and K-means clustering method, we sought to identify dynamic functional connectivity (dFC) variability patterns in both groups. The investigation into the topological structure of whole-brain functional networks utilized a graph theoretical approach. Furthermore, we employed machine learning analysis, selecting altered topological properties as classification features to distinguish RD patients from HCs. ResultsAll participants exhibited four distinct states of dynamic functional connectivity. Compared to the healthy control (HC) group, patients with RD experienced a significant reduction in the number of transitions among these four states. Additionally, the dynamic topological properties of RD patients demonstrated notable changes in both global and node-specific characteristics, with these changes correlating with clinical parameters. The support vector machine (SVM) model used for classification achieved an accuracy of 0.938, an area under the curve (AUC) of 0.988, and both sensitivity and specificity of 0.937. ConclusionThe alterations in the topological properties of the brain in RD patients may indicate the integration function and information exchange efficiency of the whole brain network were reduced. In addition, the topological properties hold considerable promise for distinguishing between RD and HCs.

Satellite Imagery, Big Data, IoT and Deep Learning Techniques for Wheat Yield Prediction in Morocco

In the domain of efficient management of resources and ensuring nutritional consistency, accuracy in predicting crop yields becomes crucial. The advancement of artificial intelligence techniques, synchronized with satellite imagery, has emerged as a potent approach for forecasting crop yields in modern times. We used two types of data: spatial data and temporal data. Spatial data are gathered from satellite imagery and processed using ArcGIS to extract data about crops based on several indices like NDVI and NWDI. Temporal data are gathered from agricultural sensors such as temperature sensors, rainfall sensor, precipitation sensor and soil moisture sensor. In our case we used Sentinel 2 satellite to extract vegetation indices. We have used IoT systems, especially Raspberry Pi B+ to collect and process data coming from sensors. All data collected are then stored into a NoSQL server to be analysed and processed. Several machine learning and deep learning algorithms have been used for the processing of crop recommendation system, such as logistic regression, KNN, decision tree, support vector machine, LSTM, and Bi-LSTM through the collected dataset. We used GRU deep learning model for the best performance, the RMSE and R2 for this model was 0,00036 and 0,99 respectively.The main contribution of our paper is the development of a new system that can predict several crop yields, such as wheat, maize, etc, using IoT, satellite imagery for spatial data and the use of sensors for temporal data. We are the first paper that has combined spatial data and temporal data to predict crop yield based on deep learning algorithms, unlike other works that uses only remote sensing data or temporal data. We created an E-monitoring crop yield prediction system that helps farmers track all information about crops and show the result of prediction in a mobile application. This system helps farmers with more efficient decision making to enhance crop production. The main production regions of wheat in Morocco are in the rainfed areas of the plains and plateaus of Chaouia, Abda, Haouz, Tadla, Gharb and Saïs. We studied three main regions well known for wheat production which are Rabat-Salé, Fez-Meknes, Casablanca-Settat.

Predicting the productivity of fractured horizontal wells using few-shot learning

Predicting the productivity of multistage fractured horizontal wells plays an important role in exploiting unconventional resources. In recent years, machine learning (ML) models have emerged as a new approach for such studies. However, the scarcity of sufficient real data for model training often leads to imprecise predictions, even though the models trained with real data better characterize geological and engineering features. To tackle this issue, we propose an ML model that can obtain reliable results even with a small amount of data samples. Our model integrates the synthetic minority oversampling technique (SMOTE) to expand the data volume, the support vector machine (SVM) for model training, and the particle swarm optimization (PSO) algorithm for optimizing hyperparameters. To enhance the model performance, we conduct feature fusion and dimensionality reduction. Additionally, we examine the influences of different sample sizes and ML models for training. The proposed model demonstrates higher prediction accuracy and generalization ability, achieving a predicted R2 value of up to 0.9 for the test set, compared to the traditional ML techniques with an R2 of 0.13. This model accurately predicts the production of fractured horizontal wells even with limited samples, supplying an efficient tool for optimizing the production of unconventional resources. Importantly, the model holds the potential applicability to address similar challenges in other fields constrained by scarce data samples.

Enhanced landslide susceptibility mapping in data-scarce regions via unsupervised few-shot learning

Given the critical need to assess landslide hazards, producing landslide susceptibility map (LSM) in regions with scarce historical landslide inventories poses significant challenges. This study introduces a novel landslide susceptibility assessment framework that combines unsupervised learning strategies with few-shot learning methods to increase the accuracy of LSM in these areas. The framework has been practically validated in a representative geological disaster-prone area along the West-East Gas Pipeline in Shaanxi Province, China. We employed three advanced few-shot learning models: a support vector machine, meta-learning, and transfer learning. These models implement feature representation learning for weakly correlated influencing factors through an unsupervised approach, thereby constructing an effective landslide susceptibility assessment model. We compared traditional learning methods and used the receiver operating characteristic (ROC) curve and SHAP values to quantify the effectiveness of the models. The results indicate that the meta-learning algorithm outperforms both the SVM and transfer learning in areas with limited landslide data. The integration of unsupervised strategies significantly improves performance, achieving area under the curve (AUC) values of 0.9385 and 0.9861, respectively. Compared with using meta-learning alone, incorporating unsupervised learning strategies increased the AUC by 4.76%, enhancing both the predictive power of the model and the interpretability of the features. Meta-learning under unsupervised conditions effectively mitigates the evaluation difficulties caused by insufficient landslide records, providing a viable path and empirical evidence for performance improvement in similar data- scarce regions worldwide.

Study on runoff forecasting and error correction driven by atmosphere–ocean-land dataset

Accurate runoff forecasting results can not only provide an important basis for flood control scheduling, but also provide scientific support for water resources optimization, which promotes the maximization of the overall benefits of the basin. To further explore the inherent mechanisms of the atmosphere–ocean-land factors driving runoff changes, this study proposes the factors dimension reduction and interpretation framework based on Pearson, eXtreme Gradient Boosting and SHapley Additive exPlanations (P-XGBoost-SHAP). Base on this, the Gaussian Process Regression (GPR), Long Short-Term Memory neural network (LSTM) and Support Vector Machine (SVM) models are used to construct the atmospheric-ocean-land data-driven runoff prediction model. Meanwhile, for the runoff prediction residuals, this paper proposes an error multi-step correction framework based on ensemble empirical mode decomposition and autoregressive model (EEMD-AR). The case study of Lianghekou hydrological station shows that the factor dimension reduction and interpretation framework can greatly reduce the input dimension of the model, and explain the factors globally and locally by using SHAP Value. Compared with the traditional Random Forest (RF) dimension reduction method, it shows higher prediction accuracy. The Nash-Sutcliffe efficiency coefficient (NSE) can be increased to about 0.93, which is 4.91 % and 1.97 % higher than the series–parallel coupling (AR-Parallel) and empirical mode decomposition-autoregressive (EMD-AR) correction methods, respectively. The accuracy of the runoff forecasting prediction is improved while reducing the input dimension of the model.

Local Neuronal Activity and the Hippocampal Functional Network Can Predict the Recovery of Consciousness in Individuals With Acute Disorders of Consciousness Caused by Neurological Injury.

There is limited research on predicting the recovery of consciousness in patients with acute disorders of consciousness (aDOC). The purpose of this study is to investigate the altered characteristics of the local neuronal activity indicated by the amplitude of low-frequency fluctuations (ALFF) and functional connectivity (FC) of the hippocampus network in patients with aDOC caused by neurological injury and to explore whether these characteristics can predict the recovery of consciousness. Thirty-seven patients with aDOC were included, all of whom completed resting-state functional magnetic resonance imaging (rsfMRI) scans. The patients were divided into two groups based on prognosis of consciousness recovery, 24 patients were in prolonged disorders of consciousness (pDOC) and 13 in emergence from minimally conscious state (eMCS) at 3 months after neurological injury. Univariable and multivariate logistic regression analyses were used to investigate the clinical indicators affecting patients' recovery of consciousness. The ALFF values and FC of the hippocampal network were compared between patients with pDOC and those with eMCS. Additionally, we employed the support vector machine (SVM) method to construct a predictive model for prognosis of consciousness based on the ALFF and FC values of the aforementioned differential brain regions. The accuracy (ACC), area under the curve (AUC), sensitivity, and specificity were used to evaluate the efficacy of the model. The FOUR score at onset and the length of mechanical ventilation (MV) were found to be significant influential factors for patients who recovered to eMCS at 3 months after onset. Patients who improved to eMCS showed significantly increased ALFF values in the right calcarine gyrus, left lingual gyrus, right middle temporal gyrus, and right precuneus compared to patients in a state of pDOC. Furthermore, significant increases in FC values of the hippocampal network were observed in the eMCS group, primarily involving the right lingual gyrus and bilateral precuneus, compared to the pDOC group. The predictive model constructed using ALFF alone or ALFF combined with FC values from the aforementioned brain regions demonstrated high accuracies of 83.78% and 81.08%, respectively, with AUCs of 95% and 94%, sensitivities of 0.92 for both models, and specificities of 0.92 for both models in predicting the recovery of consciousness in patients with aDOC. The present findings demonstrate significant differences in the local ALFF and FC values of the hippocampus network between different prognostic groups of patients with aDOC. The constructed predictive model, which incorporates ALFF and FC values, has the potential to provide valuable insights for clinical decision-making and identifying potential targets for early intervention.

Exploring EEG-Based Biomarkers for Improved Early Alzheimer's Disease Detection: A Feature-Based Approach Utilizing Machine Learning

This paper presents a comprehensive investigation into Electroencephalogram (EEG) signal processing and analysis techniques aimed at enhancing early diagnosis methods for Alzheimer's Disease (AD). Leveraging a dataset that has EEG data of individuals diagnosed with Mild Cognitive Impairment (MCI), AD, Healthy Controls, and the study explores Preprocessing Methods and Feature Extraction Techniques, with machine learning model notably Support Vector Machines (SVM).In the preprocessing phase, a combination of high pass, lowpass, Savitzky–Golay, and median filters are applied, informed by a comprehensive review of filter comparison literature. Feature extraction encompasses three primary categories: ‘Statistical, ‘Frequency Domain’ and ‘Time Domain’. The scope of this work is to explore features in all these three domains and build SVM based model for efficient classification. In our investigation, we achieved a categorization accuracy of 92% through the utilization of statistical features. Employing time domain features resulted in an accuracy of 87%, while frequency domain features also yielded an 87% accuracy rate in our study. The primary objective of this study is that it aims to enhance early AD diagnosis through advanced EEG signal processing and machine learning techniques, focusing on preprocessing methods, feature extraction, and classification accuracy.

A Machine Learning-Based Approach for Predicting the Level of Palm Oil Adulteration in Coconut Oil

Coconut oil, prized for its health benefits, faces quality threats from adulteration, particularly with cheaper palm oil. This not only degrades the quality but also poses health risks. Traditional detection methods are often labor-intensive, destructive, and time-consuming. This study addresses the issue by applying multispectral imaging technology combined with machine learning to detect palm oil adulteration in coconut oil. We selected four machine learning algorithms—Support Vector Machines (SVM), Quadratic Discriminant Analysis (QDA), K-Nearest Neighbors (KNN), and Bagging—due to their robustness in handling complex datasets. These models achieved classification accuracies of up to 100%, far surpassing traditional chemical tests, which are slower and dependent on tester expertise. To further enhance detection accuracy, we employed both hard- and soft-voting mechanisms, integrating the strengths of individual models to improve overall reliability. This research marks a significant advancement in detecting coconut oil adulteration, offering a faster, more efficient solution to ensure product quality and consumer health.

Unveiling soil coherence patterns along Etihad Rail using Sentinel-1 and Sentinel-2 data and machine learning in arid region

This research applies machine learning to predict soil coherence for Etihad Rail, marking the first comprehensive study in the United Arab Emirates (UAE)'s arid regions. By integrating Sentinel-1 SAR and Sentinel-2 data with MODIS Aerosol Optical Depth (AOD) observations, the study develops detailed models that depict complex soil coherence patterns crucial for urban planning and risk assessment. Findings show variations in soil coherence between operational and under-construction phases, influenced by seasonal changes in aerosol dynamics and sand dust levels. Higher soil coherence is linked with lower annual sand dust deposition and AOD measurements, emphasizing the importance of this data for informed decision-making. The study employs a unique combination of data sources and machine learning algorithms to predict soil coherence, including Support Vector Machine (SVM), Extreme Gradient Boosting (XGBOOST), Gaussian Process Regression (GPR), Random Forest (RF), and 1D Convolutional Neural Network (CNN), with the Random Forest model achieving the lowest root mean squared error (RMSE) of 0.0826. These contributions enhance our understanding and provide a valuable framework for infrastructure development in similar environments.

Achieving wheat seedling freezing injury assessment during the seedling stage using Unmanned Ground Vehicle (UGV) and hyperspectral imaging technology

Freezing injury may cause irreversible damage to wheat (Triticum aestivum L) tissues and can significantly reduce yield and quality. Therefore, quick and non-destructively estimating the degree of frost damage for formulating anti-freezing protection strategies and preventing frost damage is very crucial. In this study, we obtained hyperspectral images of wheat leaves for accurate identification of frost damage. A remote-controlled Unmanned Ground Vehicle (UGV) equipped with an imaging spectral camera was used to capture the hyperspectral images of frost-damaged wheat leaves. We compared the efficiency of two methods (the one without removal of weeds, and the other is to remove the corresponding area of weeds from the hyperspectral image by Deeplab V3+) for estimation of wheat freezing damage degree by using four different algorithms; Support Vector Machine Classification (SVM), Mahalanobis Distance Classification (MaD), Minimum Distance Classification (MiD), and Maximum Likelihood Classification (ML). We found that, Deeplab V3+ can efficiently identify the weeds from hyperspectral images, as the overall accuracy (OA) values of different algorithms were high in images with weeds removal as compared to the values in weeds containing images. Further, applying ML model after weeds removal have high OA (93.26 %) as compared to the other models. Thus, using Deeplab V3+ and ML can be a potential approach to identify the freezing injury in wheat for sustainable agricultural productivity.

Effective methods for quantifying complexity based on improved ordinal partition networks: Topological dispersion entropy and weighted topological dispersion entropy

Topological permutation entropy is based on ordinal partition networks (OPNs) to approximate topological entropy of low-dimensional chaotic systems. But the ordinal patterns of OPN ignore the magnitude of the amplitude value. To solve the problem, we propose topological dispersion entropy (TDE) and weighted topological dispersion entropy (WTDE) based on dispersion patterns to characterize the complexity of a system. Furthermore, WTDE strengthens the topological structure analysis of complex networks by weighting the adjacency matrix of the improved ordinal partition networks, thereby more accurately capturing the dynamic evolution of time series. The proposed methods are comprehensively evaluated by numerical experiments. The results show that the performance of both TDE and WTDE is significantly better than TPE. Especially, WTDE has good stability to parameters, data length, and noise. Finally, combining support vector machines and K-Nearest Neighbor, TDE and WTDE are applied to the classification of physiological data and mechanical failure data.

Computer Vision Foundation Models in Endoscopy: Proof of Concept in Oropharyngeal Cancer.

To evaluate the performance of vision transformer-derived image embeddings for distinguishing between normal and neoplastic tissues in the oropharynx and to investigate the potential of computer vision (CV) foundation models in medical imaging. Computational study using endoscopic frames with a focus on the application of a self-supervised vision transformer model (DINOv2) for tissue classification. High-definition endoscopic images were used to extract image patches that were then normalized and processed using the DINOv2 model to obtain embeddings. These embeddings served as input for a standard support vector machine (SVM) to classify the tissues as neoplastic or normal. The model's discriminative performance was validated using an 80-20 train-validation split. From 38 endoscopic NBI videos, 327 image patches were analyzed. The classification results in the validation cohort demonstrated high accuracy (92%) and precision (89%), with a perfect recall (100%) and an F1-score of 94%. The receiver operating characteristic (ROC) curve yielded an area under the curve (AUC) of 0.96. The use of large vision model-derived embeddings effectively differentiated between neoplastic and normal oropharyngeal tissues. This study supports the feasibility of employing CV foundation models like DINOv2 in the endoscopic evaluation of mucosal lesions, potentially augmenting diagnostic precision in Otorhinolaryngology. 4 Laryngoscope, 134:4535-4541, 2024.