Performance Of Optimization Model Research Articles

A hybrid CNN-LSTM approach for intelligent cyber intrusion detection system

As the technology is advancing more and more in the era of increasing digitalization, safeguarding networks from cyber threats is crucial. As cyber-attacks on critical infrastructure are becoming more and more sophisticated, enhancing cyber intrusion detection systems (IDS) is imperative. This paper proposes and evaluates a deep learning-based IDS using the NSL-KDD dataset, a benchmark for intrusion detection. The system pre-processes data with Recursive Feature Elimination (RFE) and a Decision Tree classifier to identify the most significant features, optimizing model performance. Various deep learning models, including ANN, LSTM, BiLSTM, CNN-LSTM, GRU, and BiGRU, have been evaluated. The CNN-LSTM model outperformed the others, with 95 % accuracy, 0.89 recall, and 0.94 f1-score. These results prove the effectiveness of the proposed IDS in accurately distinguishing between malicious and benign network traffic. Future research can explore ensemble techniques like boosting or bagging to further enhance IDS performance.

Comparative Study on Stock Movement Prediction Using Hybrid Deep Learning Model

Applying machine learning techniques in stock market prediction has evolved significantly, with deep learning methodologies gaining prominence. Conventional algorithms such as Linear Regression and Neural Networks initially dominated but struggled to capture complex temporal dependencies in financial data. Recent research has explored deep learning architectures like LSTM and CNN and hybrids such as CNN-LSTM and LSTM-CNN, showcasing promising results. However, there's a gap in research comparing these models across different datasets, particularly in predicting stock movements. This study addresses this gap by conducting a comparative analysis of deep learning and hybrid models for stock movement prediction in the Indonesian banking sector. The evaluation based on RMSE and MAE reveals that the LSTM-CNN hybrid consistently outperforms other models, showcasing its versatility and accuracy across different data characteristics. Then, exploration through hyperparameter tuning demonstrates the criticality of parameter selection in optimizing model performance. These findings contribute to advancing predictive modeling in financial markets, offering valuable insights for investors, analysts, and policymakers. Further research in hyperparameter tuning and model optimization holds promise for enhancing accuracy and reliability in stock price prediction.

MobileNet-V2: An Enhanced Skin Disease Classification by Attention and Multi-Scale Features.

The increasing prevalence of skin diseases necessitates accurate and efficient diagnostic tools. This research introduces a novel skin disease classification model leveraging advanced deep learning techniques. The proposed architecture combines the MobileNet-V2 backbone, Squeeze-and-Excitation (SE) blocks, Atrous Spatial Pyramid Pooling (ASPP), and a Channel Attention Mechanism. The model was trained on four diverse datasets such as PH2 dataset, Skin Cancer MNIST: HAM10000 dataset, DermNet. dataset, and Skin Cancer ISIC dataset. Data preprocessing techniques, including image resizing, and normalization, played a crucial role in optimizing model performance. In this paper, the MobileNet-V2 backbone is implemented to extract hierarchical features from the preprocessed dermoscopic images. The multi-scale contextual information is fused by the ASPP model for generating a feature map. The attention mechanisms contributed significantly, enhancing the extraction ability of inter-channel relationships and multi-scale contextual information for enhancing the discriminative power of the features. Finally, the output feature map is converted into probability distribution through the softmax function. The proposed model outperformed several baseline models, including traditional machine learning approaches, emphasizing its superiority in skin disease classification with 98.6% overall accuracy. Its competitive performance with state-of-the-art methods positions it as a valuable tool for assisting dermatologists in early classification. The study also identified limitations and suggested avenues for future research, emphasizing the model's potential for practical implementation in the field of dermatology.

A Probabilistic Descent Ensemble for Malware Prediction Using Deep Learning

INTRODUCTION: Introducing a Probabilistic Descent Ensemble (PDE) approach for enhancing malware prediction through deep learning leverages the power of multiple neural network models with distinct architectures and training strategies to achieve superior accuracy while minimizing false positives. OBJECTIVES: Combining Stochastic Gradient Descent (SGD) with early stopping is a potent approach to optimising deep learning model training. Early stopping, a vital component, monitors a validation metric and halts training if it stops improving or degrades, guarding against overfitting. METHODS: This synergy between SGD and early stopping creates a dynamic framework for achieving optimal model performance adaptable to diverse tasks and datasets, with potential benefits including reduced training time and enhanced generalization capabilities. RESULTS: The proposed work involves training a Gaussian NB classifier with SGD as the optimization algorithm. Gaussian NB is a probabilistic classifier that assumes the features follow a Gaussian (normal) distribution. SGD is an optimization algorithm that iteratively updates model parameters to minimize a loss function. CONCLUSION: The proposed work gives an accuracy of 99% in malware prediction and is free from overfitting and local minima.

Self-guided Knowledge-Injected Graph Neural Network for Alzheimer's Diseases.

Graph neural networks (GNNs) are proficient machine learning models in handling irregularly structured data. Nevertheless, their generic formulation falls short when applied to the analysis of brain connectomes in Alzheimer's Disease (AD), necessitating the incorporation of domain-specific knowledge to achieve optimal model performance. The integration of AD-related expertise into GNNs presents a significant challenge. Current methodologies reliant on manual design often demand substantial expertise from external domain specialists to guide the development of novel models, thereby consuming considerable time and resources. To mitigate the need for manual curation, this paper introduces a novel self-guided knowledge-infused multimodal GNN to autonomously integrate domain knowledge into the model development process. We propose to conceptualize existing domain knowledge as natural language, and devise a specialized multimodal GNN framework tailored to leverage this uncurated knowledge to direct the learning of the GNN submodule, thereby enhancing its efficacy and improving prediction interpretability. To assess the effectiveness of our framework, we compile a comprehensive literature dataset comprising recent peer-reviewed publications on AD. By integrating this literature dataset with several real-world AD datasets, our experimental results illustrate the effectiveness of the proposed method in extracting curated knowledge and offering explanations on graphs for domain-specific applications. Furthermore, our approach successfully utilizes the extracted information to enhance the performance of the GNN.

Machine learning based classification of excessive smartphone users via neuronal cue reactivity

Excessive Smartphone Use (ESU) poses a significant challenge in contemporary society, yet its recognition as a distinct disorder remains ambiguous. This study aims to address this gap by leveraging functional magnetic resonance imaging (fMRI) data and machine learning techniques to classify ESU and non-excessive smartphone users (n-ESU) based on their neural Cue-Reactivity (CR) signatures. By conducting a CR task and analyzing brain activation patterns, we identified spatial similarities between addictive smartphone use and established addictive disorders. Our approach involved employing Support Vector Machines (SVM) for classification, enhanced with feature selection methods such as Recursive Feature Elimination (RFE) and Model-based Selection and dimensionality reduction methods such as and Principal Component Analysis (PCA) to mitigate the challenges posed by limited dataset size and high dimensionality of fMRI data. The results demonstrate the effectiveness of our classification model, achieving accuracies of up to 79.9 %. Furthermore, we observed region-specific activations contributing significantly to classification accuracy, highlighting the potential biomarkers associated with ESU. External validation on longitudinal data revealed the necessity for larger training datasets to improve model generalizability. Additionally, feature selection techniques proved crucial for optimizing model performance, particularly in datasets with combined information from multiple sources. Our findings underscore the importance of incorporating more data to enhance model stability and generalizability, with implications for advancing the understanding and treatment of ESU and related disorders. Overall, our study demonstrates the promise of machine learning approaches in elucidating neural correlates of ESU and informing targeted interventions for affected individuals.

Variable selection for uncertain regression models based on elastic net method

On one hand, determining the exact number of explanatory variables in uncertain regression analysis is often challenging, as real-world scenarios are interconnected. On the other hand, it’s worth noting that even when the number of variables is determined, some variables may have minimal impact on the response variable, and retaining them would only increase the complexity of the model. Therefore, variable selection methods can help eliminate variables with minimal impact to enhance the simplicity and interpretability of the model. In light of this, this paper first introduces a novel variable selection strategy based on the elastic net method to address the variable selection problem in uncertain regression models. This method not only offers the flexibility to simplify uncertain regression analysis into Lasso estimation or Ridge estimation but also exhibits extremely high efficiency in variable selection. Secondly, to address the issue of determining tuning parameters, we introduce the cross-validation method to ensure optimal model performance. Furthermore, we propose an uncertain hypothesis testing method based on imprecise observational data, which can help us assess the suitability of estimated parameters. Finally, through a series of numerical examples, we validate the feasibility and effectiveness of this method.

Application of artificial neural network for peak load forecasting in 150 kV Semarang power system

Accurate load forecasting is essential for reliable and efficient operation of power systems. Traditional forecasting methods often struggle with capturing complex nonlinear patterns in load data. Artificial neural networks (ANNs) have emerged as a promising alternative due to their ability to learn complex relationships from historical data (Syed et al. in IEEEA 9:54992–55008, 2021. https://doi.org/10.1109/ACCESS.2021.3071654). This study investigates the potential of ANNs for short-term peak load forecasting in a 150 kV power system in Semarang, Indonesia. The study examines the impact of different input variables, including historical peak load, minimum load, population, and energy production, on forecasting accuracy. Several ANN architectures are trained and evaluated using mean absolute percentage error (MAPE) and mean squared error (MSE) metrics as reported by Demuth and De Jesús (neural network design). The results indicate that ANNs can achieve high accuracy in predicting peak load, with MAPE values below 10%. The study also demonstrates the importance of carefully selecting input variables and training parameters for optimal model performance. The findings highlight the potential of ANNs for improving load forecasting accuracy in power systems, contributing to enhanced grid reliability and operational efficiency. The findings of this study contribute to a deeper understanding of the application of ANNs in power system load forecasting. They demonstrate the potential of ANNs to achieve high accuracy and provide valuable insights into the factors influencing model performance. The findings are relevant for power system operators, researchers, and policymakers working to improve grid reliability and efficiency as reported by Prabha Kundur and Malik (Power System Stability and Control, McGraw-Hill Education, New York, 2022. https://www.accessengineeringlibrary.com/content/book/9781260473544).

Parametric multi-level modeling of Lean Six Sigma

Lean Six Sigma is a methodology that combines Lean manufacturing and Six Sigma into a single framework for process improvement. This combination merges their tools and techniques to overcome their deficiencies while achieving bottom-line improvement. However, existing literature emphasizes general LSS implementation with very little work on modeling the interaction between Leanness and Six Sigma. This work developed a multi-level model containing leanness and Six Sigma levels to generalize LSS implementation. Three hierarchical levels—Six Sigma (level 3), leanness (level 2), and a fattening level (level 1—Sigma quality level) are defined for two models (parametrized and non-parametrized). Aggregate measures (process efficiency, asset intensity, production time, etc.) are considered at levels 2 and 3 to achieve the desired optimization goal at that level. For testing, three process measures [asset intensity (AI), planned stoppages (PS), wastes and rework (WR)] from the manufacturing data of a global brand spanning 18 months were analyzed. The maximum log-likelihood method was used to estimate the model parameters and they converged to 85.2%, 11.2%, and 8.87% for AI, PS, and WR respectively, with a strong correlation between AI and PS. A metaheuristic algorithm was extended to solve both models. The obtained process yield (0.111111 and 0.4132471), defects per million opportunities (DPMO) (12,426 and 18,046), and sigma level (3.6 and 3.74) for the parametrized and non-parametrized models respectively indicated optimal model performance and a need for the strategic use of LSS tools in improving the production output. Also, further reduction in wastes, stoppages, and downtime was observed upon tuning the parametric variables. This model provides a mathematical approach to analyzing LSS implementation and provides a basis for future adaptation to any organization.

COMPARATIVE ANALYSIS OF STATE-OF-THE-ART CLASSIFIERS FOR PARKINSON'S DISEASE DIAGNOSIS

Parkinson's disease (PD) presents a growing global health challenge, with early detection being crucial for effective management and treatment. This study seeks to develop an innovative machine learning (ML) framework for the early detection of PD by integrating advanced techniques for data preprocessing, dimensionality reduction, feature selection, and ensemble classification, aiming to significantly improve detection accuracy and timeliness. The research employs a robust ML pipeline, beginning with data preprocessing using mean imputation, standardization, min-max scaling, and SMOTE (Synthetic Minority Over-sampling Technique) to handle imbalanced data. Dimensionality reduction is achieved through Principal Component Analysis (PCA), while feature selection is performed using SelectKBest coupled with the ANOVA F-test to identify the most relevant features. Four ensemble methods—Random Forest, Gradient Boosting, XGBoost, and Support Vector Machine (SVM)—are evaluated for classification. Among the classifiers tested, the Gradient Boosting model stands out with an impressive accuracy of 0.9487, demonstrating its superior performance in PD detection. Integrating multiple preprocessing, dimensionality reduction, and feature selection techniques proves essential in optimizing model performance, highlighting the importance of a multifaceted approach in handling complex datasets. This research introduces a comprehensive ML framework that combines multiple advanced techniques in a streamlined process, significantly improving the early detection of Parkinson's disease. Ensemble methods, combined with strategic feature selection and data balancing techniques, offer a novel approach that could be applied to other neurodegenerative disorders, expanding its potential impact beyond PD detection.

A Shark Inspired Ensemble Deep Learning Stacks for Ensuring the Security in Internet of Things (IoT)-Based Smart City Infrastructure

Recent advancements in the Internet of Things (IoT) have paved the way for intelligent and sustainable solutions in smart city environments. However, despite these advantages, IoT-connected devices present significant privacy and security risks, as network attacks increasingly exploit user-centric information. To protect the network against the rapidly growing number of cyber-attacks, it is essential to employ a cognitive intrusion detection system (CIDS) capable of handling complex and voluminous network data. This research presents a novel ensemble deep learning framework designed to enhance cybersecurity in IoT-based smart city ecosystems. The proposed architecture integrates Self-Attention Convolutional Neural Networks, Bidirectional Gated Recurrent Units, and Shark Smell Optimized Feed Forward Networks to create a robust, adaptive system for detecting and mitigating cyber threats. By leveraging fog computing, the model significantly reduces latency and computational overhead, making it highly suitable for large-scale IoT deployments. Extensive experimentation using the ToN-IoT dataset demonstrates the framework's exceptional performance, achieving a 99.78% detection rate across various attack types and an AUC of 0.989. The proposed model outperforms existing state-of-the-art approaches, achieving a mean fitness function value of 0.85640 and a standard deviation of 0.037630 in binary classification outcomes. In multi-class classification, the model maintains a mean fitness function value of 0.8230 and a variance of 2.28930 × 104, significantly outperforming other meta-heuristic algorithms such as Particle Swarm Optimization (PSO) and Genetic Algorithm (GA). The model exhibits superior accuracy and efficiency compared to existing state-of-the-art approaches, particularly in identifying complex and emerging threats. This research makes significant contributions by introducing innovative feature extraction techniques, optimizing model performance for resource-constrained environments, and providing a scalable solution for securing smart city infrastructure. The findings highlight the potential of ensemble deep learning approaches to fortify IoT networks against cyberattacks, paving the way for more resilient and secure smart cities.

Reinforcement learning processes as forecasters of depression remission

BackgroundAspects of reinforcement learning have been associated with specific depression symptoms and may inform the course of depressive illness. MethodsWe applied support vector machines to investigate whether blood‑oxygen-level dependent (BOLD) responses linked with neural prediction error (nPE) and neural expected value (nEV) from a probabilistic learning task could forecast depression remission. We investigated whether predictions were moderated by treatment use or symptoms. Participants included 55 individuals (n = 39 female) with a depression diagnosis at baseline; 36 of these individuals completed standard cognitive behavioral therapy and 19 were followed during naturalistic course of illness. All participants were assessed for depression diagnosis at a follow-up visit. ResultsBoth nPE and nEV classifiers forecasted remission significantly better than null classifiers. The nEV classifier performed significantly better than the nPE classifier. We found no main or interaction effects of treatment status on nPE or nEV accuracy. We found a significant interaction between nPE-forecasted remission status and anhedonia, but not for negative affect or anxious arousal, when controlling for nEV-forecasted remission status. LimitationsOur sample size, while comparable to that of other studies, limits options for maximizing and evaluating model performance. We addressed this with two standard methods for optimizing model performance (90:10 train and test scheme and bootstrapped sampling). ConclusionsResults support nEV and nPE as relevant biobehavioral signals for understanding depression outcome independent of treatment status, with nEV being stronger than nPE as a predictor of remission. Reinforcement learning variables may be useful components of an individualized medicine framework for depression healthcare.

The analysis of art design under improved convolutional neural network based on the Internet of Things technology

This work aims to explore the application of an improved convolutional neural network (CNN) combined with Internet of Things (IoT) technology in art design education and teaching. The development of IoT technology has created new opportunities for art design education, while deep learning and improved CNN models can provide more accurate and effective tools for image processing and analysis. In order to enhance the effectiveness of art design teaching and students’ creative expression, this work proposes an improved CNN model. In model construction, it increases the number of convolutional layers and neurons, and incorporates the batch normalization layer and dropout layer to enhance feature extraction capabilities and reduce overfitting. Besides, this work creates an experimental environment using IoT technology, capturing art image samples and environmental data using cameras, sensors, and other devices. In the model application phase, image samples undergo preprocessing and are input into the CNN for feature extraction. Sensor data are concatenated with image feature vectors and input into the fully connected layers to comprehensively understand the artwork. Finally, this work trains the model using techniques such as cross-entropy loss functions and L2 regularization and adjusts hyperparameters to optimize model performance. The results indicate that the improved CNN model can effectively acquire art sample data and student creative expression data, providing accurate and timely feedback and guidance for art design education and teaching, with promising applications. This work offers new insights and methods for the development of art design education.

Integration of multiscale fusion of residual neural network with 2-D gramian angular fields for lower limb movement recognition based on multi-channel sEMG signals

The human lower limb movements recognition (LLMR) plays a pivotal role in active lower limb exoskeleton robots. Employing surface electromyography (sEMG) signals for LLMR allows for the convenient, rapid and stable capture of signal variations, facilitating efficient identification of lower limb motion patterns. However, current sEMG-based LLMR methods face challenges such as incomplete feature extraction, limited contextual information and restricted feature extraction scales during feature extraction. This paper proposed a LLMR method based on Gramian Angular Fields (GAF) and multiscale fusion of Residual Neural Network (MS-ResNet). The denoised sEMG time series was transformed into Gramian Angular Difference Field (GADF) matrix based on GAF. The MS-ResNet model, incorporating ResNet and multiscale feature fusion concepts, was proposed to comprehensively capture global and local information through different-scale feature extraction and fusion, so as to improve recognition performance. sEMG signals from 11 muscles of the preferred leg of 15 healthy subjects were recorded during six common lower limb movements. Experimental analysis investigated the impact of the convolutional kernel size (k × k) in Stream 2 of MS-ResNet and the number of muscles involved on recognition performance. The study revealed that selecting k as 13, coupled with 11 muscles, yielded optimal model performance with the average cross-individual recognition accuracy reaching 97.62 %, demonstrating the model’s efficiency in LLMR. This method could provide a viable solution for developing more efficient and reliable LLMR systems, applicable to lower limb exoskeleton robots and intelligent prosthetics.

Assessment of three Artificial Intelligence models performance in the detection of bacterial spot on tomato leaves in mozambique

The agricultural sector is pivotal to Mozambique's economy through domestic consumption and exports of crops such as (Lycopersicon esculentum Mill.). However, tomato production faces several challenges, one of the most important being the bacterial diseases infestation. Artificial Intelligence (AI) has emerged as a transformative tool across industries, including agriculture, leveraging machine learning and extensive datasets to enhance productivity and sustainability. In a study conducted at Eduardo Mondlane University experimental trial station in Maputo, images of tomato leaves were collected and categorized to develop AI models for detection of bacterial spot disease. Data preprocessing included resizing and augmentation techniques to optimize model performance. Training involved Convolutional Neural Networks (CNNs) using TensorFlow and Adam optimizers. Among models tested (Resnet 50, Inception V3, VGG 16). The VGG 16 with data augmentation achieved the highest validation accuracy of 93.33% and minimal loss (0.1595%), demonstrating superior precision and generalization. This underscores the critical role of data augmentation in training robust AI models for effective Bacterial Spot detection in tomatoes, offering significant potential for improving agricultural outcomes in Mozambique.

Entity Extraction of Key Elements in 110 Police Reports Based on Large Language Models

With the rapid advancement of Internet technology and the increasing volume of police reports, relying solely on extensive human labor and traditional natural language processing methods for key element extraction has become impractical. Applying advanced technologies such as large language models to improve the effectiveness of police report extraction has become an inevitable trend in the field of police data analysis. This study addresses the characteristics of Chinese police reports and the need to extract key elements by employing large language models specific to the public security domain for entity extraction. Several lightweight (6/7b) open-source large language models were tested as base models. To enhance model performance, LoRA fine-tuning was employed, combined with data engineering approaches. A zero-shot data augmentation method based on ChatGPT and prompt engineering techniques tailored for police reports were proposed to further improve model performance. The key police report data from a certain city in 2019 were used as a sample for testing. Compared to the base models, prompt engineering improved the F1 score by approximately 3%, while fine-tuning led to an increase of 10–50% in the F1 score. After fine-tuning and comparing different base models, the Baichuan model demonstrated the best overall performance in extracting key elements from police reports. Using the data augmentation method to double the data size resulted in an additional 4% increase in the F1 score, achieving optimal model performance. Compared to the fine-tuned universal information extraction (UIE) large language model, the police report entity extraction model constructed in this study improved the F1 score for each element by approximately 5%, with a 42% improvement in the F1 score for the “organization” element. Finally, ChatGPT was employed to align the extracted entities, resulting in a high-quality entity extraction outcome.

Programming Chatbots Using Natural Language: Generating Cervical Spine MRI Impressions.

The utility of machine learning, specifically large language models (LLMs), in the medical field has gained considerable attention. However, there is a scarcity of studies that focus on the application of LLMs in generating custom subspecialty radiology impressions. The primary objective of this study is to evaluate and compare the performance of multiple LLMs in generating specialized, accurate, and clinically useful radiology impressions for degenerative cervical spine MRI reports. The study employed a comparative analysis of multiple LLMs, including OpenAI's ChatGPT-3.5 and GPT-4 (OpenAI, San Francisco, CA), Antrhopic's Claude 2 (Anthropic PBC, San Francisco, CA), Google's Bard (Google Inc., Mountain View, CA), and Meta's Llama 2 (Meta Platforms, Inc., Menlo Park, CA). This was performed during January-February 2024. These models were evaluated using a few-shot learning approach on a dataset consisting of 10 examples from 50 synthetically generated MRI reports. Performance metrics evaluated were diagnostic accuracy, stylistic accuracy, and redundancy. While Claude 2 maintained consistent high performance across 40 cases, GPT-4 required midway re-training to improve its declining scores. Both Claude 2 and GPT-4 demonstrated the ability to generate structured impressions, but Claude 2's specialized summarization capabilities provided an edge in maintaining accuracy without continuous feedback. The other LLMs' performance was subpar. The findings of this study suggest that LLMs can be a valuable tool in automating the generation of radiology impressions. Claude 2, in particular, exhibited promising results, indicating its potential for clinical implementation. However, the study also points to the necessity for further research, especially in optimizing model performance and evaluating real-world applicability.

Forward and inverse river contaminant transport modeling using group preserving scheme

Environmental concerns have necessitated the development of computational models predicting pollutant dispersion in natural water systems. Due to the ill-posed nature of the inverse contaminant transport equation, solving this equation using all stable and convergent inverse methods is impossible. Factors such as river geometry, unsteady and non-uniform flow, and tidal influences add to the complexity of the inverse problem. These factors have prompted the evaluation of Group Preserving Scheme for environmental applications. The inverse solution method derives a general equation for solving ordinary differential equations by addressing a dynamical system at negative time steps, ensuring convergence. Three test cases have been presented to evaluate Backward Group Preserving Scheme (BGPS). These have included validation using observational data from Missouri River, inverse simulation in a tidal river, and sensitivity analysis of parameters such as pollutant patterns, advection, dispersion, and decay coefficients. The dataset includes calibrated data from Missouri River, which demonstrated high accuracy, with Mean Relative Error (MRE) ranging from 2.8% to 5.0% for the inverse model. Under tidal conditions, accuracy decreases over time but remains robust, with a Nash–Sutcliffe efficiency of 0.77–0.96 and an MRE of 0.9%–5.9%. Sensitivity analysis revealed optimal model performance for Péclet numbers greater than 500. The model performed best with gradual, wide-peaked pollutant patterns and moderate decay rates (Damköhler number between 5 and 10). BGPS proves effective for transport simulation and concentration history reconstruction in complex rivers, including those with tidal influences, offering a robust tool for water contamination analysis in various flow conditions.

Enhancing Chatbot Intent Classification using Active Learning Pipeline for Optimized Data Preparation

This study presents a novel approach to enhancing chatbot intent classification through an optimized data preparation combined with Active Learning. We applied the clustering mechanism using a state-of-the-art sentence-transformers model with cosine similarity for cluster detection in order to categorize messages. This process was further refined through a dedicated Active Learning Pipeline, which focused on the most essential observations for labeling. Incorporating externally sourced labeled data from Scale AI, the labeling process was fine-tuned iteratively, until the model's performance stabilized. This approach shows promise for various datasets and tasks, suggesting a scalable solution for preparing data for supervised modeling and achieving optimal model performance in real-world commercial chatbot scenarios.

Tracking perching behavior of cage-free laying hens with deep learning technologies

Providing perches in cage-free (CF) housing offers significant benefits for laying hens, such as improved leg muscle development, bone health, reduced abdominal fat, and decreased fear and aggression. A precise detection method is essential to ensure that hens engage in perching behavior from an early age, as manual observation is often labor-intensive and sometimes inaccurate. The objectives of this study were to (1) develop and test a deep learning model for detecting perching behavior; and (2) evaluate the optimal model's performance on detecting perching behavior of laying hens of different ages. In this study, recent deep learning models, that is, YOLOv8s-PB, YOLOv8x-PB, YOLOv7-PB, and YOLOv7x-PB, were developed, trained and compared in detecting perching behavior in 4 CF rooms (200 hens/room). Perch height was up to 1.8 m from the litter floor and situated 1.5 m below the cameras. A total of 3,000 images were used, with each image featuring at least 1 hen perching. The models' detection accuracies and their performance across different age groups of hens were compared using 1-way ANOVA at a 5% significance level. The results showed that the YOLOv8x-PB model outperform all other models used, achieving the precision of 94.80%, recall of 95.10%, and mean average precision (mAP@0.50) of 97.60%. While all models proved over 94% detection precision. With optimal model, PB detection precision was highest (97.40%) for peaking phase followed by prelay (95.20%), grower (94.80%), developer (94.70%) and layers (92.70%) phases while the lowest detection precision (88.80%) was for starter phase. Detection performance was somewhat reduced by the overlapping of birds during perching and occlusion. Overall, the YOLOv8x-PB model was the most optimal in detecting perching behavior, proposing a valuable tool for CF producers to monitor the perching activities of laying hens automatically.