Performance Of Different Models Research Articles

Research on Hard Rock Pillar Stability Prediction Based on SABO-LSSVM Model

The increase in mining depth necessitates higher strength requirements for hard rock pillars, making mine pillar stability analysis crucial for pillar design and underground safety operations. To enhance the accuracy of predicting the stability state of mine pillars, a prediction model based on the subtraction-average-based optimizer (SABO) for hyperparameter optimization of the least-squares support vector machine (LSSVM) is proposed. First, by analyzing the redundancy of features in the mine pillar dataset and conducting feature selection, five parameter combinations were constructed to examine their effects on the performance of different models. Second, the SABO-LSSVM prediction model was compared vertically with classic models and horizontally with other optimized models to ensure comprehensive and objective evaluation. Finally, two data sampling methods and a combined sampling method were used to correct the bias of the optimized model for different categories of mine pillars. The results demonstrated that the SABO-LSSVM model exhibited good accuracy and comprehensive performance, thereby providing valuable insights for mine pillar stability prediction.

Machine learning approaches to predict the need forintensive care unitadmission among Iranian COVID-19 patients based on ICD-10: A cross-sectional study.

Timely identification of the patients requiring intensive care unit admission (ICU) could be life-saving. We aimed to compare different machine learning algorithms to predict the requirements for ICU admission in COVID-19 patients. We screened all patients with COVID-19 at six academic hospitals in Tehran comprising our study population. A total of 44,112 COVID-19 patients (≥18 years old) were included, among which 7722 patients were hospitalized. We used a Random Forest algorithm to select significant variables. Then, prediction models were developed using the Support Vector Machine, Naıve Bayes, logistic regression, lightGBM, decision tree, and K-Nearest Neighbor algorithms. Sensitivity, specificity, accuracy, F1 score, and receiver operating characteristic-Area Under the Curve (AUC) were used to compare the prediction performance of different models. Based on random Forest, the following predictors were selected: age, cardiac disease, cough, hypertension, diabetes, influenza & pneumonia, malignancy, and nervous system disease. Age was found to have the strongest association with ICU admission among COVID-19 patients. All six models achieved an AUC greater than 0.60. Naıve Bayes achieved the best predictive performance (AUC = 0.71). Naïve Bayes and lightGBM demonstrated promising results in predicting ICU admission needs in COVID-19 patients. Machine learning models could help quickly identify high-risk patients upon entry and reduce mortality and morbidity among COVID-19 patients.

Machine Learning Tools for Peptide Bioactivity Evaluation – Implications for Cell Culture Media Optimization and the Broader Cultivated Meat Industry

Although bioactive peptides have traditionally been studied for their health-promoting qualities in the context of nutrition and medicine, the past twenty years have seen a steady increase in their application to cell culture media optimization. Complex natural sources of bioactive peptides, such as hydrolysates, offer a sustainable and cost-effective means of promoting cellular growth, making them an essential component of scaling-up cultivated meat production. However, the sheer diversity of hydrolysates makes product selection difficult, highlighting the need for functional characterization. Traditional wet-lab techniques for isolating and estimating peptide bioactivity cannot keep pace with peptide identification using high-throughput tools such as mass spectrometry, requiring the development and use of machine learning-based classifiers.This review provides a comprehensive list of available software tools to evaluate peptide bioactivity, classified and compared based on the algorithm, training set, functionality, and limitations of the underlying models. We curated independent test sets to compare the predictive performance of different models based on specific bioactivity classification relevant to promoting cell culture growth: antioxidant and anti-inflammatory. A comprehensive screening of all bioactivity classifiers revealed that while there are approximately fifty tools to elucidate antimicrobial activity and sixteen that predict anti-inflammatory activity, fewer tools are available for other functionalities related to cell growth — five that predict antioxidant activity and two for growth factor and/or cell signaling prediction. A thorough evaluation of the available tools revealed significant issues with sensitivity, specificity, and overall accuracy. Despite the overall interest in estimating peptide bioactivity, our work highlights key gaps in the broader adoption of existing software for the specific application of cell culture media optimization in the context of cultivated meat and beyond.

Graph neural network incorporating time-varying frequency domain features with application in spatial wind speed field prediction

Precise forecasting of wind speed is vital to guarantee safe travel on bridges. Nevertheless, current research primarily concentrates on single-point prediction. When it comes to forecasting spatial wind fields, graph neural networks prove to be powerful methods, but many of them have yet to take into account the impact of frequency domain attributes. To address this, a novel model called the graph neural network with time-varying frequency features (GTF) is proposed. Its core layer employs wavelet transforms in matrix form to extract various frequency sub-series, followed by the use of adaptive adjacency matrices to identify their spatial characteristics. Furthermore, aggregating all frequency band information and exploring spatial features further through multi-kernel convolutions. The predictive performance of different models for wind speed forecasting is assessed by utilizing spatial wind field data collected on a large-span bridge. Experimental results demonstrate that spatiotemporal model GTF outperforms the other models in terms of predictive accuracy. The quantity of wind observation points and their level of correlation exert a certain impact on the GTF model, while hyperparameters such as stacking depth and wavelet decomposition parameters have minimal effects. Finally, recommended parameters for the GTF model are provided.

Adding contrast-enhanced ultrasound can improve the predictive ability of breast conventional ultrasound and mammography for pathological upgrade of biopsy-confirmed ductal carcinoma in situ

ObjectivesTo evaluate the added value of contrast-enhanced ultrasound (CEUS) on top of breast conventional imaging for predicting the upgrading of ductal carcinoma in situ (DCIS) to invasive cancer after surgery. MethodsThis retrospective study enrolled 140 biopsy-proven DCIS lesions in 138 patients and divided them into two groups based on postoperative histopathology: non-upgrade and upgrade groups. Conventional ultrasound (US), mammography (MMG), CEUS and clinicopathological (CL) features were reviewed and compared between the two groups. The predictive performance of different models (with and without CEUS features) for histologic upgrade were compared to calculate the added value of CEUS. ResultsFifty-nine (42.1 %) lesions were histologically upgraded to invasive cancer after surgery. By logistic regression analyses, we found that high-grade DCIS at biopsy (P=0.004), ultrasonographic lesion size > 20 mm (P=0.007), mass-like lesion on US (P=0.030), the presence of suspicious calcification on MMG (P=0.014), the presence of perfusion defect (P=0.005) and the area under TIC>1021.34 ml (P<0.001) on CEUS were six independent factors predicting concomitant invasive components after surgery. The CL+US+MMG model made with the four predictors in the clinicopathologic, US and MMG categories yielded an area under the receiver operating curve (AUROC) value of 0.759 (95 % CI: 0.680–0.828) in predicting histological upgrade. The combination model built by adding the two CEUS predictors to the CL+US+MMG model showed higher predictive efficacy than the CL+US+MMG model (P=0.018), as the AUROC value was improved to 0.861 (95 % CI: 0.793–0.914). ConclusionsThe addition of contrast-enhanced ultrasound to breast conventional imaging could improve the preoperative prediction of an upgrade to invasive cancer from CNB −proven DCIS lesions.

Localization of impact on CFRP structure based on fiber Bragg gratings and CNN-LSTM-Attention

Low-velocity impacts can cause microscopic and invisible damage to carbon fiber reinforced polymer (CFRP) structures, potentially compromising their integrity and leading to catastrophic failures. Therefore, obtaining precise information about the impact location is crucial for monitoring the health of CFRP structures. In this paper, an impact localization system for CFRP structures was developed by using fiber Bragg grating (FBG) sensors, and impact signals detected by FBG sensors are demodulated by edge-filtering at high speed. An impact localization method of CFRP structure based on CNN-LSTM-Attention is proposed. The time difference of arrival (TDOA) between signals from different FBG sensors are collected to characterize the impact location, and attention mechanism is introduced into the CNN-LSTM model to augment the significance of TDOA of impact signal detected by proximal FBG sensors. The model is trained using the training set, its parameters are optimized using the validation set and the localization performance of different models are compared by the test set. The proposed impact localization method based on CNN-LSTM-Attention model was verified on a CFRP plate with an experiment area of 400 mm*400 mm. Experimental results prove the effectiveness and satisfactory performance of the proposed method.

Interpretable representation learning for 3D multi-piece intracellular structures using point clouds.

A key challenge in understanding subcellular organization is quantifying interpretable measurements of intracellular structures with complex multi-piece morphologies in an objective, robust and generalizable manner. Here we introduce a morphology-appropriate representation learning framework that uses 3D rotation invariant autoencoders and point clouds. This framework is used to learn representations of complex multi-piece morphologies that are independent of orientation, compact, and easy to interpret. We apply our framework to intracellular structures with punctate morphologies (e.g. DNA replication foci) and polymorphic morphologies (e.g. nucleoli). We systematically compare our framework to image-based autoencoders across several intracellular structure datasets, including a synthetic dataset with pre-defined rules of organization. We explore the trade-offs in the performance of different models by performing multi-metric benchmarking across efficiency, generative capability, and representation expressivity metrics. We find that our framework, which embraces the underlying morphology of multi-piece structures, facilitates the unsupervised discovery of sub-clusters for each structure. We show how our approach can also be applied to phenotypic profiling using a dataset of nucleolar images following drug perturbations. We implement and provide all representation learning models using CytoDL, a python package for flexible and configurable deep learning experiments.

Fat-suppression T2 relaxation time and water fraction predict response to intravenous glucocorticoid therapy for thyroid-associated ophthalmopathy.

To investigate the value of fat-suppression (FS) T2 relaxation time (T2RT) derived from FS T2 mapping and water fraction (WF) derived from T2 IDEAL to predict the treatment response to intravenous glucocorticoids (IVGC) in patients with thyroid-associated ophthalmopathy (TAO) based on texture analysis. In this study, 89 patients clinically diagnosed with active and moderate-to-severe TAO were enroled (responsive group, 48 patients; unresponsive group, 41 patients). The baseline clinical characteristics and texture features were compared between the two groups. Multivariate analysis was performed to identify the independent predictors of treatment response to IVGC. ROC analysis and the DeLong test were used to assess and compare the predictive performance of different models. The responsive group exhibited significantly shorter disease duration and higher 90th percentile of FS T2RT and kurtosis of WF in the extraocular muscle (EOM) and 95th percentile of WF in the orbital fat (OF) than the unresponsive group. Model 2 (disease duration + WF; AUC, 0.816) and model 3 (disease duration + FS T2RT + WF; AUC, 0.823) demonstrated superior predictive efficacy compared to model 1 (disease duration + FS T2RT; AUC, 0.756), while there was no significant difference between models 2 and 3. The orbital tissues of responders exhibited more oedema and heterogeneity. Furthermore, OF is as valuable as EOM for assessing the therapeutic efficacy of IVGC. Finally, WF derived from T2 IDEAL processed by texture analysis can provide valuable information for predicting the treatment response to IVGC in patients with active and moderate-to-severe TAO. The texture features of FS T2RT and WF are different between responders and non-responders, which can be the predictive tool for treatment response to IVGC. Texture analysis can be used for predicting response to IVGC in TAO patients. TAO patients responsive to IVGC show more oedema and heterogeneity in the orbital tissues. WF from T2 IDEAL is a tool to predict the therapeutic response of TAO.

Cardiovascular magnetic resonance feature tracking derived strain analysis can predict return to training following exertional heatstroke

BackgroundExertional heatstroke (EHS) is increasingly common in young trained soldiers. However, the prognosis marker in EHS patients remains unclear. To evaluate cardiac MRI feature tracking (CMR-FT) derived left ventricle (LV) strain as a biomarker for return to training (RTT) in trained soldiers with EHS in a prospective CMR cohort. MethodsTrained soldiers (participants) with EHS underwent cardiac MR cine sequences between June 2020 and August 2023. Two-dimensional (2D) LV strain parameters were derived. At 3 months after index CMR, the participants with persistent cardiac symptoms including chest pain, dyspnea, palpitations, syncope, and recurrent heat-related illness were defined as non-RTT. Multivariable logistic regression analysis is used to develop a predictive RTT model. The performance of different models was compared using the area under curve (AUC). ResultsA total of 80 participants (median age, 21 years; interquartile range (IQR), 20-23 years) and 27 health controls (median age, 21 years; IQR, 20-22 years) were prospectively included. Of the 77 participants, 32 (41.6%) had persistent cardiac symptoms and were not able to RTT at 3 months follow-up after experiencing EHS. The 2D global longitudinal strain (GLS) was significantly impaired in EHS participants compared to the healthy control group (-15.81 ± 1.67% vs -16.93 ± 1.22%, P =.001), which also showed significantly statistical differences between participants with RTT and non-RTT (-14.99 ± 3.54% vs -16.53 ± 1.43%, P <.001). 2D-GLS (≤ -15.00%) (odds ratio, 1.53; 95% confidence interval (CI): 1.08, 2.17; P =.016) was an independent predictor for RTT even after adjusting known risk factors. 2D-GLS provided incremental prognostic value over the clinical model and conventional CMR parameters model (AUCs: 0.72 vs 0.88, P =.013; 0.79 vs 0.88, P =.023; respectively). ConclusionsTwo-dimensional global longitudinal strain (≤ -15.00%) is an incremental prognostic CMR biomarker to predict return to training in exertional heatstroke soldiers.

Multivariable linear model for predicting graft weight based on 3-dimensional volumetry in regards to body weight change of living liver donor: an observational cohort study.

The purpose of this study is to build a prediction model for estimating graft weight about different graft volumetry methods combined with other variables. Donors who underwent living-donor right hepatectomy from March 2021 to March 2023 were included. Estimated graft volume measured by conventional method and 3-dimensional (3D) software were collected as well as the actual graft weight. Linear regression was used to build a prediction model. Donor groups were divided according to the 3D volumetry of <700 cm3, 700-899 cm3, and ≥900 cm3 to compare the performance of different models. A total of 119 donors were included. Conventional volumetry showed R2 of 0.656 (P < 0.001) while 3D software showed R2 of 0.776 (P < 0.001). The R2 of the multivariable model was 0.842 (P < 0.001) including for 3D volume (β = 0.623, P < 0.001), body mass index (β = 7.648, P < 0.001), and amount of weight loss (β = -7.252, P < 0.001). The median errors between different models and actual graft weight did not differ in donor groups (<700 and 700-899 cm3), while the median error of univariable linear model using 3D software (122.5; interquartile range [IQR], 61.5-179.8) was significantly higher than multivariable-adjusted linear model (41.5; IQR, 24.8-69.8; P = 0.003) in donors with estimated graft weight ≥900 cm3. The univariable 3D volumetry model showed an acceptable outcome for donors with an estimated graft volume <900 cm3. For donors with an estimated graft volume ≥900 cm3, the multivariable-adjusted linear model showed higher accuracy.

Data-driven shear strength predictions of prestressed concrete hollow-core slabs

Although providing sufficient shear strength is crucial in the design of hollow-core slabs (HCS) to prevent shear failure, conventional shear equations for HCS could not always represent the non-linear relationship between the shear strength and influencing parameters, especially for deep HCS. Back-propagation-neural-network- (BPNN)-based models could be a possible solution since they have been widely used in predicting structural behaviour of other concrete structures. However, the feasibility of such an application has not been examined previously. Therefore, a set of BPNN-based models with and without optimisation algorithms were proposed for predicting the shear strength of prestressed HCS. The proposed BPNN-based models were trained and verified against a database of 227 HCS tested in shear. Comparisons between BPNN-based and mechanics-based models were also performed. The corresponding performances of different models were evaluated by the mean Vut/Vpre, R2, RMSE, MAE and MAPE criteria. To the author's best knowledge, this is the first study to apply hybrid-BPNN-based models for shear strength prediction of HCS. The study found that all proposed BPNN models were effective in predicting the shear strength, while the ACO-BPNN model yielded the best predictions among all the BPNNs. It was also shown that the ACO-BPNN model provided better predictions than the ACI 318 code provisions with/without the size effect factor. Besides, the well-trained ACO-BPNN model could present the size effect of slab thickness on shear strength, which was not explicitly considered in ACI 318. In summary, the ACO-BPNN-based model could a rational tool for predicting shear strength for HCS.

Vegetation Water Content Retrieval from Spaceborne GNSS-R and Multi-Source Remote Sensing Data Using Ensemble Machine Learning Methods

Vegetation water content (VWC) is a crucial parameter for evaluating vegetation growth, climate change, natural disasters such as forest fires, and drought prediction. Spaceborne global navigation satellite system reflectometry (GNSS-R) has become a valuable tool for soil moisture (SM) and biomass remote sensing (RS) due to its higher spatial resolution compared with microwave measurements. Although previous studies have confirmed the enormous potential of spaceborne GNSS-R for vegetation monitoring, the utilization of this technology to fuse multiple RS parameters to retrieve VWC is not yet mature. For this purpose, this paper constructs a local high-spatiotemporal-resolution spaceborne GNSS-R VWC retrieval model that integrates key information, such as bistatic radar cross section (BRCS), effective scattering area, CYGNSS variables, and surface auxiliary parameters based on five ensemble machine learning (ML) algorithms (i.e., bagging tree (BT), gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), random forest (RF), and light gradient boosting machine (LightGBM)). We extensively tested the performance of different models using SMAP ancillary data as validation data, and the results show that the root mean square errors (RMSEs) of the BT, XGBoost, RF, and LightGBM models in VWC retrieval are better than 0.50 kg/m2. Among them, the BT and RF models performed the best in localized VWC retrieval, with RMSE values of 0.50 kg/m2. Conversely, the XGBoost model exhibits the worst performance, with an RMSE of 0.85 kg/m2. In terms of RMSE, the RF model demonstrates improvements of 70.00%, 52.00%, and 32.00% over the XGBoost, LightGBM, and GBDT models, respectively.

Logging Lithology Discrimination with Enhanced Sampling Methods for Imbalance Sample Conditions

In the process of lithology discrimination from a conventional well logging dataset, the imbalance in sample distribution restricts the accuracy of log identification, especially in the fine-scale reservoir intervals. Enhanced sampling balances the distribution of well logging samples of multiple lithologies, which is of great significance to precise fine-scale reservoir characterization. This study employed data over-sampling and under-sampling algorithms represented by the synthetic minority over-sampling technique (SMOTE), adaptive synthetic sampling (ADASYN), and edited nearest neighbors (ENN) to process well logging dataset. To achieve automatic and precise lithology discrimination on enhanced sampled well logging dataset, support vector machine (SVM), random forest (RF), and gradient boosting decision tree (GBDT) models were trained using cross-validation and grid search methods. Aimed to objectively evaluate the performance of different models on different sampling results from multiple perspectives, the lithology discrimination results were evaluated and compared based on the Jaccard index and F1 score. By comparing the predictions of eighteen lithology discrimination workflows, a new discrimination process containing ADASYN, ENN, and RF has the most precise lithology discrimination result. This process improves the discrimination accuracy of fine-scale reservoir interval lithology, has great generalization ability, and is feasible in a variety of different geological environments.

P-685 An artificial intelligence based medication regimen recommendation model for optimizing the treatment outcomes of controlled ovarian stimulation

Abstract Study question Can a machine learning (ML) algorithm suggest an optimal controlled ovarian stimulation (COS) protocol for maximizing the number of retrieved and mature (MII) oocytes? Summary answer Based on historical electronic medical record patient data, using ML algorithm based medication regimen recommendation model may increase the number of retrieved oocytes. What is known already The application of artificial intelligence in the field of reproductive medicine is mainly focused on the image analysis of gamete and embryo, and prediction of treatment outcomes. The research on the optimization of COS process is limited. Previous studies mainly focused on one or several types of COS protocols such as GnRH-a protocol and GnRH-A protocol. And most of them aimed to assist single-node decision-making during COS process such as optimal gonadotropins starting dose and trigger day. Study design, size, duration The data used for developing this model consists of 38,355 COS cycles from women who underwent their first IVF/ICSI treatment in a large fertility center between January 2008 to December 2022, of which 33,853 were used as the model training set and 4,502 were used as the test set. This dataset contains a subset of 8,382 cycles with information about MII oocytes. Participants/materials, setting, methods The goal of model was to recommend the optimal COS protocol based on optimizing COS outcomes. The similarity of patient’s characteristic was used as model input. XGBoost algorithm was constructed to calculate the similarity of COS outcomes. The accuracy was compared with the patient similarity calculation method of KNN+Euclidean distance and KNN+cosine distance using mean absolute error (MAE). The performance of different models in improving the average number of retrieved and MII oocytes was compared. Main results and the role of chance The patient similarity calculation method based on XGBoost had lowest MAE when predicting the number of retrieved oocytes (XGBoost vs KNN+Euclidean distance+k=500 vs KNN+Euclidean distance+k=1000 vs KNN+cosine distance+k=500 vs KNN+cosine distance+k=1000: 3.729 vs 3.810 vs 3.815 vs 3.785 vs 3.788). Including AMH could improve the accuracy of all recommendation models in predicting protocols and the number of retrieved oocytes. When the optimal number of MII oocytes is used as the protocol recommended target, the patient similarity calculation method based on XGBoost had lowest MAE (XGBoost vs KNN+Euclidean distance+k=500 vs KNN+Euclidean distance+k=1000 vs KNN+cosine distance+k=500 vs KNN+cosine distance+k=1000: 2.868 vs 3.397 vs 3.452 vs 3.356 vs 3.395). The COS medication regimen recommended by the patient similarity calculation method based on COS treatment outcomes has resulted in higher numbers of retrieved oocytes in patients compared to those who did not adopt recommendation (P &amp;lt; 0.01). The medication regimen recommended model based on KNN+Euclidean distance algorithm (k = 500) has the best performance in improving the average number of retrieved oocytes（adopt recommendation vs not adopt recommendation: 13.01±6.61 vs 10.09±6.67, P &amp;lt; 0.01）There was no statistical difference in the number of MII oocytes between patients receiving the recommendation and those not adopt recommendation. Limitations, reasons for caution The medication regimen recommendation model is based solely on the predicted number of retrieved and MII oocytes, not including adverse treatment outcomes such as ovarian hyperstimulation syndrome. Moreover, the database used to train and validate the model is from a single fertility center. Wider implications of the findings The recommendation model is helpful to improve the number of retrieved oocytes. Based on the recommended COS protocol, further mining the medication content, medication time distribution and medication sequence transformation from the similar patient set can assist clinicians to develop individualized COS medication regimen for infertile patients. Trial registration number not applicable

Emotion Recognition in Lhasa Tibetan Speech based on Bi-LSTM Graph Convolutional Networks

Speech Emotion Recognition (SER) is a crucial component in the field of Human-Computer Interaction (HCI), with significant research and practical application implications. However, due to the complexity of the Tibetan language and the scarcity of datasets caused by the difficulty in collecting various dialects, there are not many research achievements in Tibetan speech recognition. Based on the foundation of constructing a TBLS1 dataset containing 6,000 Tibetan-language speech samples, an approach was devised for Tibetan speech emotion recognition. This approach leverages MFCC features and incorporates a Bi-directional Long Short-Term Memory (Bi-LSTM) network within a graph convolutional neural network. Finally, by comparing the performance of different models on this dataset, we demonstrated the feasibility of our model for Tibetan speech emotion recognition.

Pedestrian-induced lateral vibration of footbridges: A comparison study of different loading models

The pedestrian-induced lateral vibration of footbridges has attracted much attention since the London Millennium Bridge incident. Significant investigations were mainly performed by external-excited force models and self-excited force models. The first type models consider pedestrians as harmonic forces and/or spring-mass-damping (SMD) systems. The second type consider the self-excited forces of pedestrians that initialize the instability of footbridges by e.g. applying the inverted pendulum (IP) model. This study compares performance of different models in reproducing the observed results of typical living footbridges: the London Millennium Bridge (UK), T-bridge (Japan) and Pedro e Inês footbridge (Portugal). With varying step frequencies and pedestrian numbers, both the acceleration time history and the discrepancy between the input work and consumed work exhibit a consistent trend for the harmonic force (HF) model and SMD model, yet a notable difference for the IP model. The HF model and SMD model are better suited for assuming synchronization, making them convenient for estimating the serviceability of pedestrian-induced lateral vibration. On the other hand, the IP model is more appropriate for incorporating random step frequencies and can better capture observed instability phenomena realistically. Because the SMD model with (inconsistent) parameters from literature may result in significant differences in predicting the acceleration amplitudes, this paper modifies the SMD model and proposes a pedestrian synchronization ratio formula as an exponential function of the maximum lateral acceleration. It is validated by comparing the analytical and numerical results for the three footbridges. Furthermore, by the modified SMD model, the lateral lock-in phenomenon of footbridges is simulated reliably and conveniently. This work contributes to analyze pedestrian-induced lateral vibration of footbridges.

A Benchmark for Morphological Segmentation in Uyghur and Kazakh

Morphological segmentation and stemming are foundational tasks in natural language processing. They have become effective ways to alleviate data sparsity in agglutinative languages because of the nature of agglutinative language word formation. Uyghur and Kazakh, as typical agglutinative languages, have made significant progress in morphological segmentation and stemming in recent years. However, the evaluation metrics used in previous work are character-level based, which may not comprehensively reflect the performance of models in morphological segmentation or stemming. Moreover, existing methods avoid manual feature extraction, but the model’s ability to learn features is inadequate in complex scenarios, and the correlation between different features has not been considered. Consequently, these models lack representation in complex contexts, affecting their effective generalization in practical scenarios. To address these issues, this paper redefines the morphological-level evaluation metrics: F1-score and accuracy (ACC) for morphological segmentation and stemming tasks. In addition, two models are proposed for morpheme segmentation and stem extraction tasks: supervised model and unsupervised model. The supervised model learns character and contextual features simultaneously, then feature embeddings are input into a Transformer encoder to study the correlation between character and context embeddings. The last layer of the model uses a CRF or softmax layer to determine morphological boundaries. In unsupervised learning, an encoder–decoder structure introduces n-gram correlation assumptions and masked attention mechanisms, enhancing the correlation between characters within n-grams and reducing the impact of characters outside n-grams on boundaries. Finally, comprehensive comparative analyses of the performance of different models are conducted from various points of view. Experimental results demonstrate that: (1) The proposed evaluation method effectively reflects the differences in morphological segmentation and stemming for Uyghur and Kazakh; (2) Learning different features and their correlation can enhance the model’s generalization ability in complex contexts. The proposed models achieve state-of-the-art performance on Uyghur and Kazakh datasets.

Genre Classification of Books in Russian with Stylometric Features: A Case Study

Within the literary domain, genres function as fundamental organizing concepts that provide readers, publishers, and academics with a unified framework. Genres are discrete categories that are distinguished by common stylistic, thematic, and structural components. They facilitate the categorization process and improve our understanding of a wide range of literary expressions. In this paper, we introduce a new dataset for genre classification of Russian books, covering 11 literary genres. We also perform dataset evaluation for the tasks of binary and multi-class genre identification. Through extensive experimentation and analysis, we explore the effectiveness of different text representations, including stylometric features, in genre classification. Our findings clarify the challenges present in classifying Russian literature by genre, revealing insights into the performance of different models across various genres. Furthermore, we address several research questions regarding the difficulty of multi-class classification compared to binary classification, and the impact of stylometric features on classification accuracy.

Prediction value study of breast cancer tumor infiltrating lymphocyte levels based on ultrasound imaging radiomics.

Construct models based on grayscale ultrasound and radiomics and compare the efficacy of different models in preoperatively predicting the level of tumor-infiltrating lymphocytes in breast cancer. This study retrospectively collected clinical data and preoperative ultrasound images from 185 breast cancer patients confirmed by surgical pathology. Patients were randomly divided into a training set (n=111) and a testing set (n=74) using a 6:4 ratio. Based on a 10% threshold for tumor-infiltrating lymphocytes (TIL) levels, patients were classified into low-level and high-level groups. Radiomic features were extracted and selected using the training set. The evaluation included assessing the relationship between TIL levels and both radiomic features and grayscale ultrasound features. Subsequently, grayscale ultrasound models, radiomic models, and nomograms combining radiomics score (Rad-score) and grayscale ultrasound features were established. The predictive performance of different models was evaluated through receiver operating characteristic (ROC) analysis. Calibration curves assessed the fit of the nomograms, and decision curve analysis (DCA) evaluated the clinical effectiveness of the models. Univariate analyses and multivariate logistic regression analyses revealed that indistinct margin (P<0.001, Odds Ratio [OR]=0.214, 95% Confidence Interval [CI]: 0.103-1.026), posterior acoustic enhancement (P=0.027, OR=2.585, 95% CI: 1.116-5.987), and ipsilateral axillary lymph node enlargement (P=0.001, OR=4.214, 95% CI: 1.798-9.875) were independent predictive factors for high levels of TIL in breast cancer. In comparison to grayscale ultrasound model (Training set: Area under curve [AUC] 0.795; Testing set: AUC 0.720) and radiomics model (Training set: AUC 0.803; Testing set: AUC 0.759), the nomogram demonstrated superior discriminative ability on both the training (AUC 0.884) and testing (AUC 0.820) datasets. Calibration curves indicated high consistency between the nomogram model's predicted probability of breast cancer TIL levels and the actual occurrence probability. DCA revealed that the radiomics model and the nomogram model achieved higher clinical net benefits compared to the grayscale ultrasound model. The nomogram based on preoperative ultrasound radiomics features exhibits robust predictive capacity for the non-invasive evaluation of breast cancer TIL levels, potentially providing a significant basis for individualized treatment decisions in breast cancer.

Prediction of Machine-Generated Financial Tweets Using Advanced Bidirectional Encoder Representations from Transformers

With the rise of Large Language Models (LLMs), distinguishing between genuine and AI-generated content, particularly in finance, has become challenging. Previous studies have focused on binary identification of ChatGPT-generated content, overlooking other AI tools used for text regeneration. This study addresses this gap by examining various AI-regenerated content types in the finance domain. Objective: The study aims to differentiate between human-generated financial content and AI-regenerated content, specifically focusing on ChatGPT, QuillBot, and SpinBot. It constructs a dataset comprising real text and AI-regenerated text for this purpose. Contribution: This research contributes to the field by providing a dataset that includes various types of AI-regenerated financial content. It also evaluates the performance of different models, particularly highlighting the effectiveness of the Bidirectional Encoder Representations from the Transformers Base Cased model in distinguishing between these content types. Methods: The dataset is meticulously preprocessed to ensure quality and reliability. Various models, including Bidirectional Encoder Representations Base Cased, are fine-tuned and compared with traditional machine learning models using TFIDF and Word2Vec approaches. Results: The Bidirectional Encoder Representations Base Cased model outperforms other models, achieving an accuracy, precision, recall, and F1 score of 0.73, 0.73, 0.73, and 0.72 respectively, in distinguishing between real and AI-regenerated financial content. Conclusions: This study demonstrates the effectiveness of the Bidirectional Encoder Representations base model in differentiating between human-generated financial content and AI-regenerated content. It highlights the importance of considering various AI tools in identifying synthetic content, particularly in the finance domain in Pakistan.