Classification Model Research Articles

Efficient Fault Detection and Analysis of Power System Distribution Networks by Integrating BP Data Mining

As the basic guarantee for people’s production and life, the safe operation of the power system has an important impact on the development and operation of society. To ensure the safe and stable operation of the power grid, predicting potential faults and taking reasonable preventive measures can effectively avoid the occurrence of power accidents. However, due to the difficulty in ensuring the prediction accuracy of traditional methods, there are issues of protection misoperation and rejection. Therefore, in order to achieve accurate prediction of power grid faults and avoid protection misoperation and rejection issues, a distribution network fault classification prediction model using a combination of three-layer data mining model (TLDM) and adaptive moment estimation (Adam) algorithm/random gradient descent algorithm improved backpropagation neural network (BPNN) is proposed. The implementation results showed that the classification accuracy of artificial fish school [Formula: see text] priori, [Formula: see text]-means clustering convolutional neural network model and TLDM for single-phase grounding faults was 93.2%, 91.5% and 96.6%, respectively. The classification accuracy for two-phase faults was 92.8%, 92.4% and 95.7%, respectively. The classification accuracy for two-phase grounding faults was 93.7%, 91.2% and 96.9%, respectively. The classification accuracy for three-phase faults was 93.3%, 92.1% and 97.1%, respectively. The TLDM had the highest classification accuracy. The average accuracy, average accuracy and average recall of the BPNN improved by the combination of the ADAM algorithm and random gradient descent algorithm were 94.1%, 90.9% and 88%, respectively, which were higher than the BPNN improved by the combination of ADAM algorithm and random gradient descent algorithm. The above results indicate that the proposed distribution network fault classification and prediction model has good performance and can achieve accurate prediction of distribution network faults.

Deep learning approaches for classification of copper-containing metal scrap in recycling processes

Separating copper from iron scrap is a critical operation in metal recycling and achieving this with low cost sensoric equipment like RGB cameras instead of XRF/XRT is becoming increasingly attractive. In this article, the groundwork for creating an image classification model to separate copper from iron scrap has been performed. Twenty of the most common and most easily available CNN architectures were trained on 2200 metal scrap specimens and evaluated inline on a sensor-based sorting rig for their prediction accuracy and their inference latency to mimic real circumstances in an industrial setting. Out of these evaluated architectures, DenseNet-201 with 98% accuracy in inline tests is recommended if potent hardware is available. Otherwise AlexNet with 92% accuracy or MobileNet-V2 with 90% accuracy are recommended for further investigation and model creation if hardware restrictions apply. Based on the presented results in this article, the initial cumbersome surveyance of the most suitable network architecture can be substantially reduced and the creation of a sorting model can be streamlined. This article thus provides the basis for creating an inline applicable sorting method for scrap metal that uses low cost sensorics equipment and can provide reasonably high accuracy in its prediction.

Development and performance evaluation of an indoor thermal environment control algorithm incorporating MET estimation model with object detection

As people increasingly spend time indoors, the significance of the indoor environment in influencing occupant quality of life is becoming more pronounced. Traditionally, indoor environment control primarily relied on fixed temperature settings, which failed to accommodate the diverse circumstances of occupants. This approach limited the creation of a comfortable indoor environment and the enhancement of energy efficiency. Consequently, there is growing interest in occupant-centric control (OCC), which integrates metabolic rate (MET) information, which is a critical factor in determining the thermal sensation of occupants. Previously, a method was developed to estimate MET by classifying occupant poses and detecting the objects they interact with from indoor images. This study aims to develop and experimentally validate an indoor thermal environment control algorithm (ITEC-algorithm) using the MET estimation model and assess its effectiveness and applicability in real building environments.The performance evaluation revealed that the ITEC-algorithm significantly enhanced the comfort ratios, achieving improvements of up to 59% compared to the fixed temperature control and 28% compared to the control methods that only used the pose classification model for MET estimation. The energy consumption varied depending on the activity and control method, with a reduction of up to 88% compared to fixed temperature control. These results indicate that thermal comfort can be enhanced while minimizing unnecessary energy consumption by incorporating the MET of the occupants. Consequently, it has been confirmed that the ITEC-algorithm effectively improves thermal comfort by managing the MET of various occupants.

Predicting ignitability classification of thermally thick solids using hybrid GA-BPNN and PSO-BPNN algorithms

Two hybrid binary classification models combining the merits of GA (genetic algorithm) and PSO (particle swarm optimization) with BPNN (backpropagation neural network) are proposed to estimate the ignitability of thermally thick solids. A 1D numerical model verified by analytical correlation and experimental measurements is employed to yield training, validation and testing datasets. Surface absorptivity (ε), density (ρ), specific heat (Cp), thermal conductivity (k), critical temperature (Tcri), and incident heat flux (HF) which are highly related to solid ignition are selected as inputs, and ignitability serves as output. A one-hidden-layer BPNN is first constructed, and then GA and PSO are encoded to form GA-BPNN and PSO-BPNN where connection weights and biases are optimized by GA/PSO and the predicted error serves as objective function of GA/PSO. Results show that the optimal neuron number in hidden layer is 34. Compared to BPNN, GA-BPNN and PSO-BPNN feature higher accuracy but more training time. Accuracy and convergence efficiency of PSO-BPNN are higher than GA-BPNN, nevertheless GA-BPNN exhibits better robustness. Predictive accuracies of BPNN, GA-BPNN, and PSO-BPNN on testing dataset are 99.05%, 99.55%, and 99.55%, respectively. ROC (receiver operating characteristic) curves are plotted to rank the performance of the three models, revealing 0.5 is an appropriate threshold to classify ignitability. Feature importance of the models is analyzed using PI (permutation importance) and MIV (mean impact value) methods. It is found HF and Tcri are the most important two inputs followed by ε, while ρ, Cp, and k show least and approximately identical impact. Reliability of the three models is verified by predicting critical heat flux of ignition of three polymers, and good agreement with literature is observed.

GVVST: Image-Driven Style Extraction From Graph Visualizations for Visual Style Transfer.

Incorporating automatic style extraction and transfer from existing well-designed graph visualizations can significantly alleviate the designer's workload. There are many types of graph visualizations. In this paper, our work focuses on node-link diagrams. We present a novel approach to streamline the design process of graph visualizations by automatically extracting visual styles from well-designed examples and applying them to other graphs. Our formative study identifies the key styles that designers consider when crafting visualizations, categorizing them into global and local styles. Leveraging deep learning techniques such as saliency detection models and multi-label classification models, we develop end-to-end pipelines for extracting both global and local styles. Global styles focus on aspects such as color scheme and layout, while local styles are concerned with the finer details of node and edge representations. Through a user study and evaluation experiment, we demonstrate the efficacy and time-saving benefits of our method, highlighting its potential to enhance the graph visualization design process.

Automatic noise detection for ambulatory electrocardiogram in presence of ventricular arrhythmias through a machine learning approach

Noise detection in ambulatory electrocardiography is investigated as a machine learning binary classification problem on a set of twelve noise indices. Ten of these noise indices are replicated from relevant scientific literature. Two novel noise indices are also introduced: the electrocardiogram derivative pattern similarity index (edp) and the number of inversions of the signal’s slope over a minimum amplitude threshold (inv). Noise indices were computed on ex-novo labeled data, which included ventricular arrhythmias such as ventricular tachycardia, flutter, and fibrillation. Signal quality labels were assigned on windows of 2 seconds, from which a set of labels on 10-second windows was also created. Six classification models were trained on 2 s and 10 s window data with a multi-stage optimization procedure. A feature selection stage was performed for each model, probing the usefulness of different noise indices, including noise and ventricular arrhythmias. Hyperparameter selection and model assessment were conducted through nested cross-validation. Final test performance metrics reached values as high as Matthew’s correlation coefficient of 0.854 and balanced accuracy of 0.923. Particular attention was placed on the correct quality classification of records with ventricular arrhythmias, and test specificity on records with ventricular arrhythmias was as high as 0.976. This study provided new insights into the role of different noise indices in a challenging context with ventricular arrhythmias.

Ultrasound analysis of different forms of hemolytic uremic syndrome in children

BackgroundHemolytic uremic syndrome (HUS) is the most common cause of acute kidney injury in children. It is mainly caused by Shiga toxin-producing enterohemorrhagic Escherichia coli (EHEC; STEC-HUS) and is more rarely caused by uncontrolled complement activation (cHUS). Renal replacement therapy is frequently required and kidney function recovers in the majority of patients. Ultrasound (US) is the preferred imaging modality for the evaluation of any renal failure. The aim of this study is the evaluation of US diagnostics in both HUS types at disease onset and in the course of the disease.Materials and methodsClinical, laboratory, and US data from the digital patient records of children admitted as inpatients with a diagnosis of HUS were recruited for a monocentric, retrospective analysis. STEC-HUS and cHUS were diagnosed when, in addition to the laboratory constellation, EHEC infection and complement system activation were verified, respectively. US examinations were performed by pediatricians with certified pediatric US experience.ResultsIn total, 30 children with STEC-HUS (13/25 male; median age of disease onset 2.9 years; most prevalent EHEC serotype was O157) and cHUS (2/5 male; median age of disease onset 5.4 years; 3/5 with proven pathogenic variation) were included. Renal replacement therapy proportions were comparable in the STEC-HUS and cHUS patients (64% vs. 60%). The resistance index (RI) was elevated at disease onset in the patients with STEC-HUS and cHUS (0.88 ± 0.10 vs. 0.77 ± 0.04, p = 0.13) and was similar in the STEC-HUS subcohorts divided based on dialysis requirement (yes: 0.86 ± 0.1; no: 0.88 ± 0.1; p = 0.74). Total kidney size at disease onset displayed a positive correlation with dialysis duration (R = 0.53, p = 0.02) and was elevated in both HUS types (177% ± 56 and 167% ± 53). It was significantly higher in the STEC-HUS subcohort which required dialysis (200.7% vs. 145%, p < .029), and a regressor kidney size threshold value of 141% was indicated in the receiver operating characteristic analysis. A classification model using both US parameters sequentially might be of clinical use for predicting the need for dialysis in patients with STEC-HUS. The US parameters normalized over time.ConclusionThe US parameters of RI and total kidney size are valuable for the assessment of HUS at disease onset and during therapy, and may be helpful in the assessment of whether dialysis is required in patients with STEC-HUS.

A Probability-Based Model for Course Completion Prediction in Online Asynchronous Learning

The sheer scale of Massive Open Online Courses (MOOCs), presents a significant challenge in delivering personalized learning experiences and effective student support. With vast participation numbers, it becomes difficult for instructors to track individual progress, pinpoint specific areas where students struggle, and understand the underlying reasons for failure to complete the course. This lack of individualized attention can lead to disengagement and higher dropout rates. A probability-based analysis offers a solution by generating student-specific predictions about their likelihood of completion. This empowers educators to proactively identify those at risk, tailor interventions, optimize resource allocation, and potentially improve the overall learning experience within the MOOC environment. This study focuses on developing a robust predictive framework to accurately estimate the probability of students completing an introductory programming course offered on the Open Learning Platform of the University of Moratuwa, Sri Lanka. The approach began with a classification model to determine the likelihood of course completion. Building upon this, a regression model was developed to generate a specific probability percentage representing the chance of a student successfully completing the course within a designated time frame. Initial findings suggested that predictions from the classification model achieved the highest accuracy when students have completed approximately 42.8% of the course materials. It is anticipated that further refinements to the methodology will improve the reliability of the predictions. A crucial aspect of this research involves determining the optimal percentage of course progress needed to yield reliable probability predictions. This is investigated through systematic analysis and experimentations including incremental model testing. The dataset encompasses diverse demographic and educational variables, enabling the identification of influential factors affecting course completion. This study provides insights on developing personalized learning strategies, intervention tactics and academic planning within online asynchronous education.

Conceptualizing bias in EHR data: A case study in performance disparities by demographic subgroups for a pediatric obesity incidence classifier.

Electronic Health Records (EHRs) are increasingly used to develop machine learning models in predictive medicine. There has been limited research on utilizing machine learning methods to predict childhood obesity and related disparities in classifier performance among vulnerable patient subpopulations. In this work, classification models are developed to recognize pediatric obesity using temporal condition patterns obtained from patient EHR data in a U.S. study population. We trained four machine learning algorithms (Logistic Regression, Random Forest, Gradient Boosted Trees, and Neural Networks) to classify cases and controls as obesity positive or negative, and optimized hyperparameter settings through a bootstrapping methodology. To assess the classifiers for bias, we studied model performance by population subgroups then used permutation analysis to identify the most predictive features for each model and the demographic characteristics of patients with these features. Mean AUC-ROC values were consistent across classifiers, ranging from 0.72-0.80. Some evidence of bias was identified, although this was through the models performing better for minority subgroups (African Americans and patients enrolled in Medicaid). Permutation analysis revealed that patients from vulnerable population subgroups were over-represented among patients with the most predictive diagnostic patterns. We hypothesize that our models performed better on under-represented groups because the features more strongly associated with obesity were more commonly observed among minority patients. These findings highlight the complex ways that bias may arise in machine learning models and can be incorporated into future research to develop a thorough analytical approach to identify and mitigate bias that may arise from features and within EHR datasets when developing more equitable models.

A novel image classification framework based on variational quantum algorithms

Image classification is a crucial task in machine learning with widespread practical applications. The existing classical framework for image classification typically utilizes a global pooling operation at the end of the network to reduce computational complexity and mitigate overfitting. However, this operation often results in a significant loss of information, which can affect the performance of classification models. To overcome this limitation, we introduce a novel image classification framework that leverages variational quantum algorithms (VQAs) hybrid approaches combining quantum and classical computing paradigms within quantum machine learning. The major advantage of our framework is the elimination of the need for the global pooling operation at the end of the network. In this way, our approach preserves more discriminative features and fine-grained details in the images, which enhances classification performance. Additionally, employing VQAs enables our framework to have fewer parameters than the classical framework, even in the absence of global pooling, which makes it more advantageous in preventing overfitting. We apply our method to different state-of-the-art image classification models and demonstrate the superiority of the proposed quantum architecture over its classical counterpart through a series of state vector simulation experiments on public datasets. Our experiments show that the proposed quantum framework achieves up to a 9.21% increase in accuracy and up to a 15.79% improvement in F1 score, compared to the classical framework. Additionally, we explore the impact of shot noise on our method through shot-based simulation and find that increasing the number of measurements does not always lead to better results. Selecting an appropriate number of measurements can yield optimal results, even surpassing those obtained from state vector simulation.

Decoding Sustainability Reporting: Insights from Text Mining and Classification Models for SDGs

Objective: This study aims to investigate the signaling effects of Sustainability Reporting (SR) in listed Singapore firms, particularly in demonstrating firm capitalization and sustainability performance. The research focuses on identifying distinct SR signals using advanced text mining and classification techniques, and their alignment with the United Nations' Sustainable Development Goals (SDGs). Theoretical Framework: The study is grounded in signaling theory, which posits that firms communicate observable characteristics (SR content) to signal unobservable attributes (firm capabilities and sustainability performance). By applying this framework, we analyze how SR may act as a signaling mechanism for large-cap firms and sustainability leaders. Method: This research collects 1,844 sustainability reports from listed Singapore firms between 2016 and 2022. Using a combination of Structural Topic Modeling (STM) and Random Forest (RF) classification models, we extracted and analyzed the key SR signals that differentiate firms by size and sustainability performance. Results: The results indicate that SR signals such as community engagement, climate risk, and compliance are effective in distinguishing between large firms and firms with strong sustainability performance. The Random Forest models achieved high accuracy in identifying these groups. Community engagement, compliance with the Global Reporting Initiative (GRI), and sustainable construction were among the most impactful topics, with large firms focusing more on scope and governance, while sustainability leaders emphasized environmental risks and compliance. Research Implications: The findings provide insights into the practical and theoretical implications of SR as a signaling tool. Key SR topics are identified as critical signals for firms with large capitalization and high sustainability performance, guiding firms in utilizing SR in the context of SDGs. Originality/Value: The findings show that SR topics such as community engagement, climate risk, and compliance serve as key signals for differentiating firms by size and sustainability performance. Firms can strategically emphasize these areas to enhance their SR practices while contributing to broader SDGs.The findings provide insights into the practical and theoretical implications of SR as a signaling tool. Key SR topics are identified as critical signals for firms with large capitalization and high sustainability performance, guiding firms in enhancing their SR strategies.

A Framework for Detecting Pulmonary Diseases from Lung Sound Signals Using a Hybrid Multi-Task Autoencoder-SVM Model

Research focuses on the efficacy of Multi-Task Autoencoder (MTAE) models in signal classification due to their ability to handle many tasks while improving feature extraction. However, researchers have not thoroughly investigated the study of lung sounds (LSs) for pulmonary disease detection. This paper introduces a new framework that utilizes an MTAE model to detect lung diseases based on LS signals. The model integrates an autoencoder and a supervised classifier, simultaneously optimizing both classification accuracy and signal reconstruction. Furthermore, we propose a hybrid approach that combines an MTAE and a Support Vector Machine (MTAE-SVM) to enhance performance. We evaluated our model using LS signals from a publicly available database from King Abdullah University Hospital. The model attained an accuracy of 89.47% for four classes (normal, pneumonia, asthma, and chronic obstructive pulmonary disease) and 90.22% for three classes (normal, pneumonia, and asthma cases). Using the MTAE-SVM, the accuracy was further improved to 91.49% for four classes and 93.08% for three classes, respectively. The results indicate that the MTAE and MTAE-SVM have a considerable potential for detecting pulmonary diseases from lung sound signals. This could aid in the creation of more user-friendly and effective diagnostic tools.

Fuzzy Soft Sets and its Application to Decision Making: A Short Case Study Involving the Health Sector

The health sector faces uncertainty and complex decision-making scenarios, making traditional analytical tools insufficient. The fuzzy soft set theory has emerged as a powerful framework for modeling and reasoning with uncertain information, with promising applications in the health domain. This project explores the application of fuzzy soft sets in various decision-making processes in the health sector, including medical diagnosis, disease classification, treatment planning, risk assessment, patient stratification, and predictive modeling. The study reviews historical development of fuzzy set theory and its extension to soft sets, discussing challenges, limitations, and future research directions. The findings aim to contribute to the growing body of knowledge on the practical relevance and potential of fuzzy soft set theory in addressing healthcare decision-making needs.

Research on Abnormal State Detection of CZ Silicon Single Crystal Based on Multimodal Fusion

The Czochralski method is the primary technique for single-crystal silicon production. However, anomalous states such as crystal loss, twisting, swinging, and squareness frequently occur during crystal growth, adversely affecting product quality and production efficiency. To address this challenge, we propose an enhanced multimodal fusion classification model for detecting and categorizing these four anomalous states. Our model initially transforms one-dimensional signals (diameter, temperature, and pulling speed) into time–frequency domain images via continuous wavelet transform. These images are then processed using a Dense-ECA-SwinTransformer network for feature extraction. Concurrently, meniscus images and inter-frame difference images are obtained from the growth system’s meniscus video feed. These visual inputs are fused at the channel level and subsequently processed through a ConvNeXt network for feature extraction. Finally, the time–frequency domain features are combined with the meniscus image features and fed into fully connected layers for multi-class classification. The experimental results show that the method can effectively detect various abnormal states, help the staff to make a more accurate judgment, and formulate a personalized treatment plan for the abnormal state, which can improve the production efficiency, save production resources, and protect the extraction equipment.

Graph neural network-based subgraph analysis for predicting adverse drug events

PurposeAdverse drug events (ADEs) are a significant global public health concern, and they have resulted in high rates of hospital admissions, morbidity, and mortality. Prior to the use of machine learning and deep learning methods, ADEs may not become well recognized until long after a drug has been approved and is widely used, which poses a significant challenge for ensuring patient safety. Consequently, there is a need to develop computational approaches for earlier identification of ADEs not detected during pre-registration clinical trials. MethodsThis paper presents a state-of-the-art network-based approach that models patients as subgraphs composed of nodes of International Classification of Diseases (ICD) codes and directed edges illustrating disease progression. Four Graph Neural Network (GNN) variants were employed to make sub-graph level predictions that answer three Research Questions (RQ): 1) whether ADE(s) would occur given a patient's prior diagnoses history, 2) when an ADE would occur, and 3) which ADE would occur. The first and second RQs were addressed using a binary classification approach. The third RQ was addressed using a multi-label classification model. ResultsThe proposed network-based approach demonstrated superior performance in predicting ADEs, with the GraphSage model exhibiting the highest accuracy for both RQ 1 (0.8863) and RQ 3 (0.9367), while the Graph Attention Networks (GAT) model was found to perform best for RQ 2 (0.8769). Furthermore, an analysis segmented by ADE classification revealed that while RQs 1 and 3 exhibited minimal variance across different ADE categories, a distinct advantage was observed for categories B, C, and E in the context of RQ 2 when applying this sub-graph method. ConclusionThe network-based approach demonstrates the potential of GNNs in supporting the early detection and prevention of ADEs. Accurately predicting ADEs could enable healthcare professionals to make informed clinical decisions, take preventive measures and adjust medication regimens before serious adverse events occur. The proposed prediction method could also lead to optimized usage of healthcare resources by preventing hospital admissions and reducing the overall burden of adverse drug events on the healthcare systems.

A Feature-Fusion Technique-Based Alzheimer’s Disease Classification Using Magnetic Resonance Imaging

Background: Early identification of Alzheimer’s disease (AD) is essential for optimal treatment and management. Deep learning (DL) technologies, including convolutional neural networks (CNNs) and vision transformers (ViTs) can provide promising outcomes in AD diagnosis. However, these technologies lack model interpretability and demand substantial computational resources, causing challenges in the resource-constrained environment. Hybrid ViTs can outperform individual ViTs by visualizing key features with limited computational power. This synergy enhances feature extraction and promotes model interpretability. Objectives: Thus, the authors present an innovative model for classifying AD using MRI images with limited computational resources. Methods: The authors improved the AD feature-extraction process by modifying the existing ViTs. A CatBoost-based classifier was used to classify the extracted features into multiple classes. Results: The proposed model was generalized using the OASIS dataset. The model obtained an exceptional classification accuracy of 98.8% with a minimal loss of 0.12. Conclusions: The findings highlight the potential of the proposed AD classification model in providing an interpretable and resource-efficient solution for healthcare centers. To improve model robustness and applicability, subsequent research can include genetic and clinical data.

Enhancing COVID-19 disease severity classification through advanced transfer learning techniques and optimal weight initialization schemes

ObjectiveMedical imaging plays a central role in medicine today by supporting diagnosis and treatment. For small medical image datasets, training from scratch is a time-consuming process and transfer learning emerges as a solution. In these cases, ImageNet weights are usually used as initial weights, after which fine-tuning is performed. We propose a methodology for COVID-19 severity classification (mild, moderate, severe, critical) in chest X-ray images based on transfer learning using DenseNet121 architecture as a base model. MethodsThe novelty of the approach mainly lies in the investigation of three different weight initialization schemes (i) ImageNet (ii) CheXNeXt (iii) DeepCOVID-XR, which are similar in a graded manner in terms of image nature, with ImageNet being the least similar, CheXNeXt similar as they were trained on different lung diseases not including COVID-19 and DeepCOVID-XR the most similar as they were obtained by training specifically on COVID-19 X-ray images. ResultsThe results show that the worst results are achieved using ImageNet as initial weights (average AUC = 0.700), followed by the better results with DeepCOVID-XR (average AUC = 0.774) while CheXNeXt performed significantly better (average AUC = 0.917). The results of weaker performing classification models were improved when the severe and critical classes were merged, to account for the similarity between these classes in the dataset (average AUC for initialization schemes were ImageNet AUC = 0.867, CheXNeXt AUC = 0.900 and DeepCOVID-XR AUC = 0.794). ConclusionThis leads to a conclusion that, in the medical domain, where image datasets are usually small and highly imbalanced, if initial weights are chosen to be in nature similar to the new dataset, achieved results are better. However, there is no need to start from weights obtained during training on the same disease, as this may cause overfitting. There are significant discrepancies between ImageNet and medical imaging datasets and the results from this paper could help in guiding future implementations of transfer learning in medical applications.

Identification of drought-salinity combined stress in tomato plants by vegetation indices

A major issue in several farming areas of the Mediterranean basin consists of drought and salinity stress. This stress is mainly due to a steady exposition of warm daily temperature and heatwaves, moreover with inevitable irrigation with saline water. Therefore, detecting the stress is essential to minimise significant yield loss and preserve agricultural sustainability. In this context, remote and proximal sensing can play a crucial role in allowing fast, not destructive, extensive, and reliable assessment of crop status. In this work, the effectiveness of several multispectral indices in detecting salinity and water stress in tomato plants, grown under controlled green-house conditions, was investigated. Three different classifiers (fine tree model, linear discriminant model, and linear support vector machines model) were used to verify whether, and the extent to which, the adopted multispectral indices can be adopted to identify a stress condition of the tomato plants. In the experimental campaign, the stress occurrence on tomato plants was assessed on the base of a set of ecophysiological measurements, such as transpiration, stomatal conductance, and photosynthesis rate. Obtained results showed that a classification model based on linear support vector machines, exploiting the combination of Photochemical Reflectance Index and the Chlorophyl Index, can detect drought and salinity stress in tomato plants with an accuracy higher than 94%.

Deep Learning Models for Anatomical Location Classification in Esophagogastroduodenoscopy Images and Videos: A Quantitative Evaluation with Clinical Data

Background/Objectives: During gastroscopy, accurately identifying the anatomical locations of the gastrointestinal tract is crucial for developing diagnostic aids, such as lesion localization and blind spot alerts. Methods: This study utilized a dataset of 31,403 still images from 1000 patients with normal findings to annotate the anatomical locations within the images and develop a classification model. The model was then applied to videos of 20 esophagogastroduodenoscopy procedures, where it was validated for real-time location prediction. To address instability of predictions caused by independent frame-by-frame assessment, we implemented a hard-voting-based post-processing algorithm that aggregates results from seven consecutive frames, improving the overall accuracy. Results: Among the tested models, InceptionV3 demonstrated superior performance for still images, achieving an F1 score of 79.79%, precision of 80.57%, and recall of 80.08%. For video data, the InceptionResNetV2 model performed best, achieving an F1 score of 61.37%, precision of 73.08%, and recall of 57.21%. These results indicate that the deep learning models not only achieved high accuracy in position recognition for still images but also performed well on video data. Additionally, the post-processing algorithm effectively stabilized the predictions, highlighting its potential for real-time endoscopic applications. Conclusions: This study demonstrates the feasibility of predicting the gastrointestinal tract locations during gastroscopy and suggests a promising path for the development of advanced diagnostic aids to assist clinicians. Furthermore, the location information generated by this model can be leveraged in future technologies, such as automated report generation and supporting follow-up examinations for patients.

Comparison Between Gans and Diffusion Models in the Generation of Synthetic Images for Enhancing Tree Species Recognition

Objective: This study investigates the application of Generative Adversarial Networks (GANs) and Diffusion Models in data augmentation to improve the classification of tree species images, which is essential for sustainable forest management. Theoretical Framework: Fundamental concepts of machine learning, generative and classifier networks, as well as data augmentation techniques through synthetic image generation, are presented, establishing a solid foundation for the research. Method: The research utilized 2,178 images of cross-sections of wood from 18 species, applying Deep Convolutional GAN (DCGAN) and U-Net with diffusion on a rare species, evaluated using FID and IS metrics. The generated images were used to train and validate classification models, assessed by F1-score. Results and Discussion: The results revealed that Diffusion Models generated more realistic images and performed better in the classification of the rare species. The discussion contextualizes these results in light of the theoretical framework, exploring their implications for environmental management. Limitations, such as the impact of indiscriminate addition of synthetic data, are also addressed. Research Implications: The practical and theoretical implications of this research are discussed, providing insights into how the results can be applied or influence practices in the field of forest management and environmental management. These implications may cover areas such as sustainable innovation and biodiversity conservation. Originality/Value: This study contributes to the literature by demonstrating the potential of Diffusion Models to outperform GANs in generating synthetic images for sustainable forest management purposes. The relevance and value of this research are evidenced by its practical application in promoting innovative and sustainable practices in environmental management.