Performance Of Different Models Research Articles

Multi-ontology embeddings approach on human-aligned multi-ontologies representation for gene-disease associations prediction

ObjectivesKnowledge graphs and ontologies in the biomedical domain provide rich contextual knowledge for a variety of challenges. Employing that for knowledge-driven NLP tasks such as gene-disease association prediction represents a promising way to increase the predictive power of a model. MethodsWe investigated the power of infusing the embedding of two aligned ontologies as prior knowledge to the NLP models. We evaluated the performance of different models on some large-scale gene-disease association datasets and compared it with a model without incorporating contextualized knowledge (BERT). ResultsThe experiments demonstrated that the knowledge-infused model slightly outperforms BERT by creating a small number of bridges. Thus, indicating that incorporating cross-references across ontologies can enhance the performance of base models without the need for more complex and costly training. However, further research is needed to explore the generalizability of the model. We expected that adding more bridges would bring further improvement based on the trend we observed in the experiments. In addition, the use of state-of-the-art knowledge graph embedding methods on a joint graph from connecting OGG and DOID with bridges also yielded promising results. ConclusionOur work shows that allowing language models to leverage structured knowledge from ontologies does come with clear advantages in the performance. Besides, the annotation stage brought out in this paper is constrained in reasonable complexity.

Mineral resource assessment through geostatistical analysis in a phosphate deposit

Purpose. The selection of an appropriate variographic model is crucial in geostatistics to obtain accurate estimates of mineral reserves. The aim of this work is to develop a reserve estimation tool using a geostatistical approach. Methodology. The geostatistical approach is based on selecting the most representative variographic models for the studied variables. The model selection is done by applying a cross-validation procedure leave-one-out (LOOCV). LOOCV is a resampling technique used in statistical analysis and machine learning to estimate the generalization error of a model and compare the performance of different models. The studied variables are then estimated using ordinary kriging. Findings. The application of the proposed approach has resulted in satisfactory results in terms of dispersion of grades and thicknesses of mineralized layers in a phosphate deposit. To evaluate the quality of the adjustment models obtained, efficiency factors such as Nash-Sutcliffe, and RMSE (Root Mean Square Error), were employed. These factors provide quantitative measures of the agreement between the observed and predicted values. The NSE (Nash-Sutcliffe efficiency) and RMSE (root mean square error) values of 0.572 and 6.599, respectively, indicate a better fit and greater accuracy of the adjustment models. The accuracy and efficiency criteria of the studied variables have acceptable values, with a mean square error (MSE) of 1.54 · 10-7. Originality. The combination of the least squares and LOOCV methods in the geostatistical analysis leads to improved estimation precision, greater reliability in representing the spatial variability of the parameters, and enhanced confidence in the validity of the adjustment models. Practical value. The development of a computer code for this geostatistical approach provides a practical tool for decision-makers to use in the management and exploitation of mining sites. Overall, this study has contributed to the advancement of geostatistical techniques and their application in the mining industry.

Arabic Toxic Tweet Classification: Leveraging the AraBERT Model

Social media platforms have become the primary means of communication and information sharing, facilitating interactive exchanges among users. Unfortunately, these platforms also witness the dissemination of inappropriate and toxic content, including hate speech and insults. While significant efforts have been made to classify toxic content in the English language, the same level of attention has not been given to Arabic texts. This study addresses this gap by constructing a standardized Arabic dataset specifically designed for toxic tweet classification. The dataset is annotated automatically using Google’s Perspective API and the expertise of three native Arabic speakers and linguists. To evaluate the performance of different models, we conduct a series of experiments using seven models: long short-term memory (LSTM), bidirectional LSTM, a convolutional neural network, a gated recurrent unit (GRU), bidirectional GRU, multilingual bidirectional encoder representations from transformers, and AraBERT. Additionally, we employ word embedding techniques. Our experimental findings demonstrate that the fine-tuned AraBERT model surpasses the performance of other models, achieving an impressive accuracy of 0.9960. Notably, this accuracy value outperforms similar approaches reported in recent literature. This study represents a significant advancement in Arabic toxic tweet classification, shedding light on the importance of addressing toxicity in social media platforms while considering diverse languages and cultures.

A systematic study of the performance of machine learning models on analyzing the association between semen quality and environmental pollutants

Human exposure to Phthalates, a family of chemicals primarily used to enhance the flexibility and durability of plastics, could lead to a decline in semen quality. Extensive studies have been performed to investigate the associations between semen quality and exposure to environmental pollutants, such as phthalates. However, these early studies mainly focus on using conventional statistical methods, such as simple and efficient multi-variable linear regression methods, to perform the analysis, which may not be effective in analyzing these complex multi-variable associations. Herein, we perform a systematic study of the performance of different machine learning methods in analyzing these associations. We will use data from a cohort of 1070 Chinese males from Hubei province who provided repeated urine samples to measure phthalate metabolites. In addition, phthalate metabolites in semen are also evaluated as a biomarker to give a more direct metric. We also incorporate patient demographics and administered medications into the analysis. Overall, six machine learning models, including linear and non-linear models, are implemented to analyze associations among thirty-one features and five metrics of the quality of the semen. The performance of the models is evaluated based on root-mean-square deviation through 10-fold cross-validation. Our investigations show that the performance of different models is varied when employed to study different metrics that represent the quality of the semen. Therefore, a systematic study of the patients’ data with various machine learning models is essential in improving the quantitative analysis in discovering the critical environmental pollutants that dictate the quality of semen. We hope this study could provide guidance of employing machine learning models in the future investigation of the impact of various pollutants on semen quality.

Deep learning prediction of waterflooding-based alteration of reservoir hydraulic flow unit

The hydraulic flow unit (HFU) is a comprehensive representation of reservoir quality, and the type of reservoir HFU alters dramatically due to the impact of waterflooding. The accurate prediction of HFU characteristic alterations based-waterflooding is crucial but difficult because the alterations are influenced by many factors like geological conditions and production history. In this study, we propose a deep learning-based method using both static geological data and dynamic production history data to predict HFU characteristic alterations waterflooding-based. We classify six HFU types using the k-means clustering of geological variables, including the flow zone indicator (FZI), permeability, and shale volume. The deep-learning model uses the initial static geological variables, including FZI, permeability, and shale volume, and dynamic production data as inputs, whereas the HFU type after waterflooding is used as the target. Convolutional neural networks (CNNs) are adopted for the prediction of HFU characteristic alterations and also compared with support vector machines (SVM) and back propagation neural networks (BPNNs). The prediction accuracy of the CNN model is 93.3% and 87.1% for the validation and testing datasets, respectively, which performs much better than other machine learning methods. In addition, we explore the factors affecting the performance of different models. CNN model can extract useful information from complicated production history data more effectively than other machine learning methods. Finally, the trained CNN model is used to predict the spatial distribution of HFU at different development stages in the study area. In summary, the proposed method can accurately predict dynamic alterations in the reservoir HFU type after waterflooding, which is very useful for the adjustment of oilfield production measures.

A survey of few-shot knowledge graph completion

With the continuous development of knowledge graph completion (KGC) technology, the problem of few-shot knowledge graph completion (FKGC) is becoming increasingly prominent. Traditional methods for KGC are not effective in addressing this problem due to the lack of sufficient data samples. Therefore, completing the task of knowledge graph with few-shot data has become an urgent issue that needs to be addressed and solved. This paper first presents a concise introduction to FKGC, which covers relevant definitions and highlights the advantages of FKGC techniques. We then categorize FKGC methods into meta-learning-based, metric-based, and graph neural network-based methods, and analyze the unique characteristics of each model. We also introduced the research on FKGC in a specific domain - Temporal Knowledge Graph Completion (TKGC). Subsequently, we summarized the commonly used datasets and evaluation metrics in existing methods and evaluated the completion performance of different models in TKGC. Finally, we presented the challenges faced by FKGC and provided directions for future research.

From Digitization and Images to Text and Content: Transkribus as a Case Study

ABSTRACTThis poster explores the potential of using technological tools, specifically the Transkribus platform, for the transcription of Hebrew manuscripts. The digitization of historical resources has made them accessible, but the textual content of the scanned images remains inaccessible. Transkribus, an AI‐powered platform, offers tools for text recognition, transcription, and search of historical documents. The poster discusses the process of automatic text recognition (ATR) and the challenges it faces, particularly in handling handwritten texts and Hebrew letters. It provides an overview of the Transkribus platform, its functionalities, and the training process for creating transcription models. The author presents a case study of transcribing a 15th‐century Sephardic semi‐cursive Hebrew manuscript using the Transkribus platform and evaluates the performance of different models. The poster concludes by discussing the implications and possibilities of using Transkribus for automatic transcription of historical Hebrew manuscripts. While the results show promising improvements in accuracy, further challenges and solutions are also discussed. Overall, Transkribus offers significant potential for the study and transcription of Hebrew manuscripts, revolutionizing the field of Jewish studies and historical research.

An Improved Data-Driven Method for Steering Feedback Torque of Driving Simulator

Traditional modeling methods of steering feedback torque (SFT) widely used in driving simulator are mainly physics-based, but suffers from high model complexity and low model accuracy due to the inaccurate or unknown structure, dynamics, and related parameters of the steering system, such as backlash or dry friction. For these problems, data-driven methods were considered to have advantageous in modeling such a complex yet dynamic system. However, due to the limitation of the data collected for model training from real vehicle test, the commonly used artificial neural network (ANN) cannot fully reflect the steering dynamics, and does not work well in many cases where collected data do not cover, and thus lead to poor generalization performance. In this article, a long short-term memory (LSTM) model is proposed and proved to simulate well on the dynamics of SFT, and works well in most driving conditions. A driving simulator is developed to verify the performance of different models in terms of generalization of working conditions. Based on the driving simulator data, we found that ANN failure generally arises in the transition between conventional and unconventional operating conditions, which is recognized by the proposed driving pattern recognition method. The LSTM-based modeling method can effectively solve this problem. The comparative analysis reveals that the improved LSTM model can simulate the SFT with higher accuracy and broader adaptability to driving simulator conditions.

Predictive Analysis of Students’ Learning Performance Using Data Mining Techniques: A Comparative Study of Feature Selection Methods

The utilization of data mining techniques for the prompt prediction of academic success has gained significant importance in the current era. There is an increasing interest in utilizing these methodologies to forecast the academic performance of students, thereby facilitating educators to intervene and furnish suitable assistance when required. The purpose of this study was to determine the optimal methods for feature engineering and selection in the context of regression and classification tasks. This study compared the Boruta algorithm and Lasso regression for regression, and Recursive Feature Elimination (RFE) and Random Forest Importance (RFI) for classification. According to the findings, Gradient Boost for the regression part of this study had the least Mean Absolute Error (MAE) and Root-Mean-Square Error (RMSE) of 12.93 and 18.28, respectively, in the case of the Boruta selection method. In contrast, RFI was found to be the superior classification method, yielding an accuracy rate of 78% in the classification part. This research emphasized the significance of employing appropriate feature engineering and selection methodologies to enhance the efficacy of machine learning algorithms. Using a diverse set of machine learning techniques, this study analyzed the OULA dataset, focusing on both feature engineering and selection. Our approach was to systematically compare the performance of different models, leading to insights about the most effective strategies for predicting student success.

Added value of CE-CT radiomics to predict high Ki-67 expression in hepatocellular carcinoma

BackgroundThis study aimed to develop a computed tomography (CT) model to predict Ki-67 expression in hepatocellular carcinoma (HCC) and to examine the added value of radiomics to clinico-radiological features.MethodsA total of 208 patients (training set, n = 120; internal test set, n = 51; external validation set, n = 37) with pathologically confirmed HCC who underwent contrast-enhanced CT (CE-CT) within 1 month before surgery were retrospectively included from January 2014 to September 2021. Radiomics features were extracted and selected from three phases of CE-CT images, least absolute shrinkage and selection operator regression (LASSO) was used to select features, and the rad-score was calculated. CE-CT imaging and clinical features were selected using univariate and multivariate analyses, respectively. Three prediction models, including clinic-radiologic (CR) model, rad-score (R) model, and clinic-radiologic-radiomic (CRR) model, were developed and validated using logistic regression analysis. The performance of different models for predicting Ki-67 expression was evaluated using the area under the receiver operating characteristic curve (AUROC) and decision curve analysis (DCA).ResultsHCCs with high Ki-67 expression were more likely to have high serum α-fetoprotein levels (P = 0.041, odds ratio [OR] 2.54, 95% confidence interval [CI]: 1.04–6.21), non-rim arterial phase hyperenhancement (P = 0.001, OR 15.13, 95% CI 2.87–79.76), portal vein tumor thrombus (P = 0.035, OR 3.19, 95% CI: 1.08–9.37), and two-trait predictor of venous invasion (P = 0.026, OR 14.04, 95% CI: 1.39–144.32). The CR model achieved relatively good and stable performance compared with the R model (AUC, 0.805 [95% CI: 0.683–0.926] vs. 0.678 [95% CI: 0.536–0.839], P = 0.211; and 0.805 [95% CI: 0.657–0.953] vs. 0.667 [95% CI: 0.495–0.839], P = 0.135) in the internal and external validation sets. After combining the CR model with the R model, the AUC of the CRR model increased to 0.903 (95% CI: 0.849–0.956) in the training set, which was significantly higher than that of the CR model (P = 0.0148). However, no significant differences were found between the CRR and CR models in the internal and external validation sets (P = 0.264 and P = 0.084, respectively).ConclusionsPreoperative models based on clinical and CE-CT imaging features can be used to predict HCC with high Ki-67 expression accurately. However, radiomics cannot provide added value.

Estimation of Obesity Levels through the Proposed Predictive Approach Based on Physical Activity and Nutritional Habits.

Obesity is the excessive accumulation of adipose tissue in the body that leads to health risks. The study aimed to classify obesity levels using a tree-based machine-learning approach considering physical activity and nutritional habits. Methods: The current study employed an observational design, collecting data from a public dataset via a web-based survey to assess eating habits and physical activity levels. The data included gender, age, height, weight, family history of being overweight, dietary patterns, physical activity frequency, and more. Data preprocessing involved addressing class imbalance using Synthetic Minority Over-sampling TEchnique-Nominal Continuous (SMOTE-NC) and feature selection using Recursive Feature Elimination (RFE). Three classification algorithms (logistic regression (LR), random forest (RF), and Extreme Gradient Boosting (XGBoost)) were used for obesity level prediction, and Bayesian optimization was employed for hyperparameter tuning. The performance of different models was evaluated using metrics such as accuracy, recall, precision, F1-score, area under the curve (AUC), and precision-recall curve. The LR model showed the best performance across most metrics, followed by RF and XGBoost. Feature selection improved the performance of LR and RF models, while XGBoost's performance was mixed. The study contributes to the understanding of obesity classification using machine-learning techniques based on physical activity and nutritional habits. The LR model demonstrated the most robust performance, and feature selection was shown to enhance model efficiency. The findings underscore the importance of considering both physical activity and nutritional habits in addressing the obesity epidemic.

MRI Radiomics-Based Machine Learning Models for Ki67 Expression and Gleason Grade Group Prediction in Prostate Cancer.

The Ki67 index and the Gleason grade group (GGG) are vital prognostic indicators of prostate cancer (PCa). This study investigated the value of biparametric magnetic resonance imaging (bpMRI) radiomics feature-based machine learning (ML) models in predicting the Ki67 index and GGG of PCa. A total of 122 patients with pathologically proven PCa who had undergone preoperative MRI were retrospectively included. Radiomics features were extracted from T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), and apparent diffusion coefficient (ADC) maps. Then, recursive feature elimination (RFE) was applied to remove redundant features. ML models for predicting Ki67 expression and GGG were constructed based on bpMRI and different algorithms, including logistic regression (LR), support vector machine (SVM), random forest (RF), and K-nearest neighbor (KNN). The performances of different models were evaluated with receiver operating characteristic (ROC) analysis. In addition, a joint analysis of Ki67 expression and GGG was performed by assessing their Spearman correlation and calculating the diagnostic accuracy for both indices. The ML model based on LR and ADC + T2 (LR_ADC + T2, AUC = 0.8882) performed best in predicting Ki67 expression, and ADC_wavelet-LHH_firstorder_Maximum had the highest feature weighting. The SVM_DWI + T2 (AUC = 0.9248) performed best in predicting GGG, and DWI_wavelet HLL_glcm_SumAverage had the highest feature weighting. The Ki67 and GGG exhibited a weak positive correlation (r = 0.382, p < 0.001), and LR_ADC + DWI had the highest diagnostic accuracy in predicting both (0.6230). The proposed ML models are suitable for predicting both Ki67 expression and GGG in PCa. This algorithm could be used to identify indolent or invasive PCa with a noninvasive, repeatable, and accurate diagnostic method.

A comprehensive comparison of copula models and multivariate normal distribution for geo-material parametric data

Multivariate distribution models for geo-material parametric data play a crucial role in reliability-based design. Currently, two major models are commonly used: the conventional multivariate normal distribution model and the copula model (including subtypes like the elliptical copula model and Vine copula model). This study aims to assess the performance of different models including three specific copula models (Gaussian copula, t copula, and D-Vine copula models) and the conventional normal distribution model. To this end, this study employs an incomplete database comprising 693 data samples related to lacustrine soils near Dianchi Lake in Kunming city, China. Results indicate that copula models generally exhibit lower model uncertainty compared to the conventional multivariate normal distribution model. Using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), the elliptical copula ranks first, followed by Gaussian copula, with D-Vine copula ranking third. For a balanced evaluation of performance and complexity, the conventional multivariate normal distribution model is favored. This study provides valuable insights for selecting appropriate models for geo-material parametric data, aiding reliability-based design in geotechnical engineering. Additionally, it is demonstrated that sampling with replacement in the Bootstrapping technique has only a slight effect on the estimation of correlation matrices for the incomplete database.

A model including standardized weight improved predicting waiting list mortality in adolescent liver transplant candidates: A US national study.

The Model for End-Stage Liver Disease (MELD) score has been employed to identify adolescents eligible for liver transplantation since 2004. However, the optimal model for prioritizing adolescent candidates is uncertain. In our study, we aimed at evaluating the value of adding anthropometric variables to liver transplantation allocation models among adolescents. We conducted a retrospective cohort study using the data from the Organ Procurement and Transplantation Network Standard Transplant Analysis and Research to identify adolescent patients registered on the liver transplant waiting list in the United States between January 1, 2003, and December 31, 2022. Adolescents (12-17 y) who were listed for their first liver transplantation were included. We evaluated the performance of different models including pediatric end-stage liver disease with Na and creatinine, MELD, and MELD 3.0. Furthermore, we evaluated whether adding anthropometric variables ( z -score for weight and height) would improve the models' performance for our primary outcome (mortality at 90 days after listing). We identified 1421 eligible adolescent patients. Adding a z -score of weight (MELD-TEEN) improved the performance and discrimination of the MELD score. The final model including weight z -score (MELD-TEEN) had better discriminative power compared to MELD 3.0 and pediatric end-stage liver disease with Na and creatinine in the overall cohort and in different age groups (ages 12-14 and 15-17). MELD-TEEN could improve the accuracy of allocation of liver transplants among adolescents by incorporating the weight z -score compared to MELD 3.0 and pediatric end-stage liver disease with Na and creatinine.

Scoring epidemiological forecasts on transformed scales.

Forecast evaluation is essential for the development of predictive epidemic models and can inform their use for public health decision-making. Common scores to evaluate epidemiological forecasts are the Continuous Ranked Probability Score (CRPS) and the Weighted Interval Score (WIS), which can be seen as measures of the absolute distance between the forecast distribution and the observation. However, applying these scores directly to predicted and observed incidence counts may not be the most appropriate due to the exponential nature of epidemic processes and the varying magnitudes of observed values across space and time. In this paper, we argue that transforming counts before applying scores such as the CRPS or WIS can effectively mitigate these difficulties and yield epidemiologically meaningful and easily interpretable results. Using the CRPS on log-transformed values as an example, we list three attractive properties: Firstly, it can be interpreted as a probabilistic version of a relative error. Secondly, it reflects how well models predicted the time-varying epidemic growth rate. And lastly, using arguments on variance-stabilizing transformations, it can be shown that under the assumption of a quadratic mean-variance relationship, the logarithmic transformation leads to expected CRPS values which are independent of the order of magnitude of the predicted quantity. Applying a transformation of log(x + 1) to data and forecasts from the European COVID-19 Forecast Hub, we find that it changes model rankings regardless of stratification by forecast date, location or target types. Situations in which models missed the beginning of upward swings are more strongly emphasised while failing to predict a downturn following a peak is less severely penalised when scoring transformed forecasts as opposed to untransformed ones. We conclude that appropriate transformations, of which the natural logarithm is only one particularly attractive option, should be considered when assessing the performance of different models in the context of infectious disease incidence.

Development and validation of convolutional neural network-based model to predict the risk of sentinel or non-sentinel lymph node metastasis in patients with breast cancer: a machine learning study

For patients with sentinel lymph node (SLN) metastasis and low risk of residual non-SLN (NSLN) metastasis, axillary lymph node (ALN) dissection could lead to overtreatment. This study aimed to develop and validate an automated preoperative deep learning-based tool to predict the risk of SLN and NSLN metastasis in patients with breast cancer (BC) using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) images. In this machine learning study, we retrospectively enrolled 988 women with BC from three hospitals in Zhejiang, China between June 1, 2013 to December 31, 2021, June 1, 2017 to December 31, 2021, and January 1, 2019 to June 30, 2023, respectively. Patients were divided into the training set (n=519), internal validation set (n=129), external test set 1 (n=296), and external test set 2 (n=44). A convolutional neural network (CNN) model was proposed to predict the SLN and NSLN metastasis and was compared with clinical and radiomics approaches. The performance of different models to detect ALN metastasis was measured by the area under the curve (AUC), accuracy, sensitivity, and specificity. This study is registered at ChiCTR, ChiCTR2300070740. For SLN prediction, the top-performing model (i.e., the CNN algorithm) achieved encouraging predictive performance in the internal validation set (AUC 0.899, 95% CI, 0.887-0.911), external test set 1 (AUC 0.885, 95% CI, 0.867-0.903), and external test set 2 (AUC 0.768, 95% CI, 0.738-0.798). For NSLN prediction, the CNN-based model also exhibited satisfactory performance in the internal validation set (AUC 0.800, 95% CI, 0.783-0.817), external test set 1 (AUC 0.763, 95% CI, 0.732-0.794), and external test set 2 (AUC 0.728, 95% CI, 0.719-0.738). Based on the subgroup analysis, the CNN model performed well in tumour group smaller than 2.0cm, with the AUC of 0.801 (internal validation set) and 0.823 (external test set 1). Of 469 patients with BC, the false positive rate of SLN prediction declined from 77.9% to 32.9% using CNN model. The CNN model can predict the SLN status of any detectable lesion size and condition of NSLN in patients with BC. Overall, the CNN model, employing ready DCE-MRI images could serve as a potential technique to assist surgeons in the personalized axillary treatment of in patients with BC non-invasively. National Key Research and Development projects intergovernmental cooperation in science and technology of China, National Natural Science Foundation of China, Natural Science Foundation of Zhejiang Province, and Zhejiang Medical and Health Science Project.

Using CT imaging features to predict visceral pleural invasion of non-small-cell lung cancer

To examine the diagnostic performance of different models based on computed tomography (CT) imaging features in differentiating the invasiveness of non-small-cell lung cancer (NSCLC) with multiple pleural contact types. A total of 1,573 patients with NSCLC (tumour size ≤3 cm) were included retrospectively. The clinical and pathological data and preoperative imaging features of these patients were investigated and their relationships with visceral pleural invasion (VPI) were compared statistically. Multivariate logistic regression was used to eliminate confounding factors and establish different predictive models. By univariate analysis and multivariable adjustment, surgical history, tumour marker (TM), number of pleural tags, length of solid contact and obstructive inflammation were identified as independent risk predictors of pleural invasiveness (p=0.014, 0.003, <0.001, <0.001, and 0.017, respectively). In the training group, comparison of the diagnostic efficacy between the combined model including these five independent predictors and the image feature model involving the latter three imaging predictors were as follows: sensitivity of 88.9% versus 77% and specificity of 73.5% versus 84.1%, with AUC of 0.868 (95% CI: 0.848-0.886) versus 0.862 (95% CI: 0.842-0.880; p=0.377). In the validation group, the sensitivity and specificity of these two models were as follow: the combined model, 93.5% and 74.3%, the imaging feature model, 77.4% and 81.3%, and their areas under the curve (AUCs) were both 0.884 (95% CI: 0.842-0.919). The best cut-off value of length of solid contact was 7.5 mm (sensitivity 68.9%, specificity 75.5%). The image feature model showed great potential in predicting pleural invasiveness, and had comparable diagnostic efficacy compared with the combined model containing clinical data.

Generating Chinese Event Extraction Method Based on ChatGPT and Prompt Learning

Regarding the scarcity of annotated data for existing event extraction tasks and the insufficient semantic mining of event extraction models in the Chinese domain, this paper proposes a generative joint event extraction model to improve existing models in two aspects. Firstly, it utilizes the content generation capability of ChatGPT to generate annotated data corpora for event extraction tasks and trains the model using supervised learning methods adapted to downstream tasks. Secondly, explicit entity markers and event knowledge are added to the text to construct generative input templates, enhancing the performance of event extraction. To validate the performance of this model, experiments are conducted on DuEE1.0 and Title2Event public datasets, and the results show that both data enhancement and prompt learning based on ChatGPT effectively improve the performance of the event extraction model, and the F1 values of the events extracted by the CPEE model proposed in this paper reach 85.1% and 59.9% on the two datasets, respectively, which are comparable to the existing models’ values of 1.3% and 10%, respectively; moreover, on the Title2Event dataset, the performance of different models on the event extraction task can be gradually improved as the data size of the annotated corpus of event extraction generated using ChatGPT increases.

Performance Evaluation of UAVs in Wheat Disease Control

Plant protection unmanned aircraft vehicles (UAVs) were developed rapidly in China. The operation performances of different models of UAVs were different. This paper systematically studied droplet deposition distribution; pesticide-mixture utilization rate; operational efficiency; wheat-disease control efficacy; and a comprehensive score of three types of UAVs, a boom sprayer, and a knapsack sprayer. The results showed the descending order of the droplet penetration rate of the pesticide application equipment (PAE) was boom sprayer; UAVs; and knapsack sprayer. The pesticide-mixture utilization rates of the UAVs and boom sprayer were more than 50% while that of the knapsack electric sprayer was only 27.8%. The UAVs’ average labor productivity was 5.75 ha per man-hour, which was slightly less than that of the boom sprayer and 21.3 times that of the knapsack sprayer. The control efficacy of each machine on wheat Fusarium head blight was more than 90%. The average performance comprehensive score of the UAVs was 0.812, which was slightly lower than the score of 0.929 for the 3WPZ-700 self-propelled boom sprayer but much higher than the score of 0.399 for the 3WBD-18 knapsack electric sprayer. The results clearly showed the potential of UAVs for improving the pesticide-mixture utilization rate and operational efficiency, as well as the wheat Fusarium head blight control efficacy.

When Object Detection Meets Knowledge Distillation: A Survey.

Object detection (OD) is a crucial computer vision task that has seen the development of many algorithms and models over the years. While the performance of current OD models has improved, they have also become more complex, making them impractical for industry applications due to their large parameter size. To tackle this problem, knowledge distillation (KD) technology was proposed in 2015 for image classification and subsequently extended to other visual tasks due to its ability to transfer knowledge learned by complex teacher models to lightweight student models. This paper presents a comprehensive survey of KD-based OD models developed in recent years, with the aim of providing researchers with an overview of recent progress in the field. We conduct an in-depth analysis of existing works, highlighting their advantages and limitations, and explore future research directions to inspire the design of models for related tasks. We summarize the basic principles of designing KD-based OD models, describe related KD-based OD tasks, including performance improvements for lightweight models, catastrophic forgetting in incremental OD, small object detection, and weakly/semi-supervised OD. We also analyze novel distillation techniques, i.e. different types of distillation loss, feature interaction between teacher and student models, etc. Additionally, we provide an overview of the extended applications of KD-based OD models on specific datasets, such as remote sensing images and 3D point cloud datasets. We compare and analyze the performance of different models on several common datasets and discuss promising directions for solving specific OD problems.