Machine Model Research Articles

Evaluating machine learning models for predictive analytics of liver disease detection using healthcare big data

Liver diseases rank among the most prevalent health issues globally, causing significant morbidity and mortality. Early detection of liver diseases allows for timely intervention, which can prevent the progression of such diseases to more severe stages such as cirrhosis or liver cancer. To this end, many machine learning models have been previously developed to early predict liver diseases among potential patients. However, each model has its accuracy and performance limitations. In this paper, we present a comprehensive comparison of three different machine learning models that can be employed to enhance the prediction and management of liver diseases. We utilize a big data set of 32,000 records to evaluate the performance of each model. First, we implement a preprocessing technique to rectify missing or corrupt data in liver disease datasets, ensuring data integrity. Afterwards, we compare the performance of three machine models: k-nearest neighbors (KNN), gaussian naive Bayes (Gaussian NB) and random forest (RF). We concluded that the RF algorithm demonstrates superior performance in our evaluation, excelling in both predictive accuracy and the ability to classify patients accurately regarding the presence of liver disease. Our results show that RF outperforms other models based on several performance metrics including accuracy: 97.3%, precision: 97%, recall: 96%, and f1-score: 95%.

Interpretable prediction of acute ischemic stroke after hip fracture in patients 65 years and older based on machine learning and SHAP

BackgroundHip fracture and acute ischemic stroke (AIS) are prevalent conditions among the older population. The prognosis for older patients who experience AIS subsequent to hip fracture is frequently unfavorable. MethodsPatients were categorized into the AIS group and the non-AIS group. A predictive model was developed using six different machine learning algorithms. The SHapley Additive exPlanations (SHAP) method was then utilized to provide both local and global explanations. We performed adjusted mediation analyses. Furthermore, a nomogram was created to present the outcomes obtained from the LASSO regression examination. The main objective was to ascertain influential elements that can predict the occurrence of AIS. To alleviate the influence of confounding variables, propensity score matching was utilized to compare the occurrence of additional complications. Survival was compared by Kaplan-Meier methods. ResultsThe AUC of 6 ML models ranged from 0.73 to 0.87. The SVM model exhibited the greatest efficacy in forecasting AIS among older individuals with hip fractures. The leading 6 variables in the support vector machines (SVM) model were identified as systemic inflammatory response index (SIRI), carotid atherosclerosis, prior stroke, C-reactive protein (CRP), fibrinogen (FIB), and hypertension. The leading 2 variables in SHAP were identified as FIB at admission and SIRI index. There wasn't potential mediating effect of admission FIB between the SIRI index and AIS. There were statistically significant differences between the two groups in survival (P=0.003). ConclusionsThe model displayed good performance for prediction of AIS after hip fracture in patients 65 years and older, which might facilitate to establishment of a better clinical assessment plan.

Analysing DNA methylation and transcriptomic signatures to predict prostate cancer recurrence risk.

Prostate cancer (PCa) remains a significant global health challenge, with approximately 1.6 million new cases and 366,000 deaths annually. Despite high survival rates for localized prostate cancer, recurrence poses a substantial risk due to inherent biological factors and residual disease. Early detection and intervention are essential for enhancing patient outcomes and reducing mortality. However, traditional diagnostics such as PSA tests, digital rectal examinations, and biopsies often lack specificity resulting in overdiagnosis. There is a pressing need for novel biomarkers to enhance precision medicine approaches for PCa. This study employs a machine learning approach to identify DNA methylation and RNA expression biomarkers predictive of PCa recurrence using datasets from The Cancer Genome Atlas (TCGA). We analyzed 49,133 genes, identifying 684 differentially methylated genes (DMGs) and 691 differentially expressed genes (DEGs) between recurrence and non-recurrence groups. Ten genes (TNNI2, SPIN2, COL5A3, RNF169, CCND1, FGFR1, SLC17A2, FAMM71F2, RREB1, AOX1) were found to have significant correlations between methylation and expression, forming the basis for our predictive model. A support vector machine (SVM) model was developed using these ten genes, achieving an area under the curve (AUC) of 0.773, demonstrating robust predictive capability. Multivariate regression analysis confirmed the SVM score as an independent predictor of recurrence (HR = 0.45; 95% CI 0.28-0.69, P < 0.001). The analysis of recurrence-free survival suggested that patients with low-risk scores experienced significantly better outcomes compared to those with high-risk scores. Functional enrichment analyses of DMGs revealed significant involvement in biological processes such as transcription regulation, signal transduction, and immune response, highlighting the potential mechanistic pathways of these biomarkers. Validation using real-time PCR confirmed differential expression and methylation patterns of the identified genes in prostate cancer (PC3) and non-cancerous cell lines (PNT2). In conclusion, our study hihglights the DNA methylation biomarkers linked to PCa recurrence and introduces a promising SVM model for early prediction, potentially improving treatment outcomes. Further research is needed to explore the biological roles of these genes in PCa aiming to refine therapeutic approaches.

Evaluation of wheat drought resistance using hyperspectral and chlorophyll fluorescence imaging.

Photosynthesis drives crop growth and production, and strongly affects grain yields; therefore, it is an ideal trait for wheat drought resistance breeding. However, studies of the negative effects of drought stress on wheat photosynthesis rates have lacked accurate evaluation methods, as well as high-throughput techniques. We investigated photosynthetic capacity under drought stress in wheat varieties with varying degrees of drought stress resistance using hyperspectral and chlorophyll fluorescence (ChlF) imaging data. We analyzed various morpho-physiological traits involved in wheat drought tolerance, including tiller number, leaf relative water content, and malondialdehyde content, to determine the relationships between drought resistance and hyperspectral and ChlF data. The results showed that the spectral first derivative ratio (FDR) between drought stress and control conditions in the 680-760nm region was closely related to photosynthetic capacity and drought tolerance and that hyperspectral imaging can be used to monitor ChlF parameters, with bands sensitive to ChlF identified in two spectral regions (539-764nm and 832-989nm). The spectral first derivative at 989nm had the strongest linear relationship with the minimal fluorescence (R2=0.49). An uninformative variable elimination algorithm indicated that FDRs in the green (504-609nm), red (724-751nm), and near-infrared (944-946nm) light regions had great potential as indices of drought resistance. A support vector machine model based on the FDRs of these characteristic bands identified wheat drought resistance with 97.33% accuracy. These findings provide insight into the application of high-throughput technologies in studying drought resistance and photosynthesis in wheat.

Mathematical modeling of water sorption isotherms in specialty coffee beans processed by wet and semidry postharvest methods

This study investigates the experimental assessment and mathematical modeling of the water sorption isotherms in dried specialty coffee beans processed by wet and semidry postharvest methods. The wet and semidry sorption isotherms were experimentally obtained over a range of water activities between 0.1 and 0.85 at temperatures of 25, 35, and 45 °C using the dynamic dew point method (DDI). Mathematical modeling was conducted to describe the influence of water activity, temperature, and postharvest method on the equilibrium moisture content. Twelve conventional sorption equations and four machine learning techniques were employed for modeling, using 75% of the experimental data for training and 25% for validation. The selection of the best model was carried out via multifactor Analysis of Variance (ANOVA). Experimental results showed that wet and semidry coffee beans exhibited a type II S-shaped isotherm (Brunauer–Emmett–Teller classification) and a significant (p < 0.05) influence of temperature on sorption curves. Additionally, the mucilaginous coating found in semidry coffee beans provided a protective role against water sorption. The Support Vector Machine (SVM) model provided the best fit for describing the sorption isotherms (mean relative error, MRE < 1% and adjusted coefficient of determination, R2adj > 99%), demonstrating its robustness in predicting the equilibrium moisture content as a function of water activity, temperature, and postharvest processing method. This mathematical model could serve as a virtual representation of the storage process, facilitating real-time decision-making to enhance coffee quality management during storage.

Complexity of scheduling few types of jobs on related and unrelated machines

Abstract The task of scheduling jobs to machines while minimizing the total makespan, the sum of weighted completion times, or a norm of the load vector are among the oldest and most fundamental tasks in combinatorial optimization. Since all of these problems are in general -hard, much attention has been given to the regime where there is only a small number k of job types, but possibly the number of jobs n is large; this is the few job types, high-multiplicity regime. Despite many positive results, the hardness boundary of this regime was not understood until now. We show that makespan minimization on uniformly related machines ( $$Q|HM|C_{\max }$$ Q | H M | C max ) is -hard already with 6 job types, and that the related Cutting Stock problem is -hard already with 8 item types. For the more general unrelated machines model ( $$R|HM|C_{\max }$$ R | H M | C max ), we show that if the largest job size $$p_{\max }$$ p max or the number of jobs n is polynomially bounded in the instance size |I|, there are algorithms with complexity $$|I|^{{{\,\mathrm{\textrm{poly}}\,}}(k)}$$ | I | poly ( k ) . Our main result is that this is unlikely to be improved because $$Q||C_{\max }$$ Q | | C max is $$\mathsf {W[1]}$$ W [ 1 ] -hard parameterized by k already when n, $$p_{\max }$$ p max , and the numbers describing the machine speeds are polynomial in |I|; the same holds for $$R||C_{\max }$$ R | | C max (without machine speeds) when the job sizes matrix has rank 2. Our positive and negative results also extend to the objectives $$\ell _2$$ ℓ 2 -norm minimization of the load vector and, partially, sum of weighted completion times $$\sum w_j C_j$$ ∑ w j C j . Along the way, we answer affirmatively the question whether makespan minimization on identical machines ( $$P||C_{\max }$$ P | | C max ) is fixed-parameter tractable parameterized by k, extending our understanding of this fundamental problem. Together with our hardness results for $$Q||C_{\max }$$ Q | | C max , this implies that the complexity of $$P|HM|C_{\max }$$ P | H M | C max is the only remaining open case.

Clustering-Based Hybrid Model for Predicting Symptoms for Colorectal Cancer: A Fuzzy Machine Learning Approach

Colorectal cancer (CRC) is a form of cancer that originates in the colon (large intestine) or rectum, which are components of the digestive system. It generally begins as small, benign growths known as polyps that can form on the inner lining of the colon or rectum. In Malaysia, colorectal cancer ranks among the most prevalent cancers, especially within the Chinese and Malay communities. A study released by the Malaysian National Cancer Registry indicates that colorectal cancer is the second most common cancer type following breast cancer, impacting both genders. The occurrence of colorectal cancer in Malaysia has been consistently increasing, with approximately 15–20% of cancer cases in the nation being colorectal cancer. This type of cancer arises when cells in the body start to multiply excessively, leading to various symptoms. In this research, a novel hybrid fuzzy linear regression with symmetric parameter clustering combined with a support vector machine (FLRWSPCSVM) model is employed to predict the high-risk symptoms associated with the onset of colorectal cancer in Malaysia. The study analyzed secondary data from 180 patients diagnosed with colorectal cancer and treated in a general hospital in Kuala Lumpur, considering twenty-five independent variables with various combinations of variable types. Furthermore, the model included parameters, errors, and explanations, along with two statistical measurement errors. The findings revealed that FLRWSPCSVM identified ovarian symptoms and a history of cancer symptoms as high-risk indicators for the development of colorectal cancer, with the lowest mean square error (MSE) recorded at 100.605 and a root mean square error (RMSE) of 10.030.

Machine learning for predicting pavement roughness and optimising maintenance

This study examines four machine learning algorithms for predicting short-term and long-term International Roughness Index (IRI) changes in pavement performance. Using over 2,700 samples from the Long-Term Pavement Performance database, models are developed to forecast IRI changes under various maintenance scenarios. The approach accounts for immediate IRI changes caused by treatments and includes separate models for predicting weather and traffic variables. Random Forest models demonstrated the best performance, achieving R² values of 80–96% for long-term IRI changes and 85–92% for short-term changes across different maintenance strategies. Subsequently, Support Vector Machine (SVM) models with radial, linear, and polynomial kernels presented R² values as high as 70–84%, 60–92%, and 60–83%, respectively. The methodology enables accurate prediction of pavement roughness, facilitating treatment prioritisation and offering time and cost savings in data collection. It provides a comprehensive schedule for treatment execution, including types, timing, locations, and costs, thereby enhancing pavement management efficiency.

Synergistic effect of artificial intelligence and new real-time disassembly sensors: Overcoming limitations and expanding application scope

Abstract With the acceleration of industrialization and the increase in the complexity of equipment, it is particularly important to accurately predict and effectively maintain equipment failures. In response to the problem of difficulty in accurately achieving predictive maintenance and the severely limited application range of new real-time disassembly sensors, this article conducts research on the synergistic effect of the new real-time disassembly sensors and artificial intelligence technology. First, relevant parameter data of a certain enterprise’s generator can be collected on-site, and the Pearson correlation coefficient method can be used to calculate the correlation between the generator data parameters and the target fault type, ensuring the degree of correlation in extracting data features. Then, based on the gated recurrent unit (GRU) model, the article applied the particle swarm optimization (PSO) algorithm to optimize the parameters of the GRU network and used the support vector machine (SVM) model to optimize the classification function of the network output. Finally, the optimized GRU model can be applied to predict the fault types of generators, and based on this, it can be applied to the energy industry, agriculture, medical and health, and transportation industries to verify the application scalability of artificial intelligence and new real-time disassembly sensors. The experimental results show that the optimized GRU model combined with the new real-time disassembly sensor achieved an average accuracy of 0.94 in generator fault prediction, with a cost loss rate of only 3.03%, a decrease of 5.91% compared to the single new real-time disassembly sensor. The combination of artificial intelligence technology for precise predictive maintenance of generators greatly reduces maintenance costs and overcomes the limitations of individual real-time sensor disassembly. This has to some extent expanded the application scope and promoted the intelligent development of various industries.

Prediction of Color Change in Heat-Treated Wood Based on Improved Zebra Algorithm Optimized Deep Hybrid Kernel Extreme Learning Machine Model (IZOA-DHKELM)

In this study, an Improved Zebra Optimization Algorithm (ZOA) is proposed based on the search mechanism of the Sparrow Optimization Algorithm (SSA), the perturbation mechanism of the Particle Swarm Algorithm (PSO), and the adaptive function. Then, Improved Zebra Optimization Algorithm (IZOA) was used to optimize the Deep Hybrid Kernel Extreme Learning Machine Model (DHKELM), and the IZOA-DHKELM was obtained. The model has been used to predict the color of heat-treated wood for different species, temperatures, times, media, and profile types. In this article, the original DHKELM and the ZOA-DHKELM were compared to verify the validity and accuracy of the model. The results indicated that the IZOA-DHKELM decreased the mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) by 56.2%, 67.4%, and 34.2%, respectively, while enhancing the coefficient of determination, R2, to 0.9952 compared to the ZOA-DHKELM. This demonstrated that the model was significantly optimized, with improved generalization ability and prediction accuracy. It can better meet the actual engineering needs.

A copy number variation detection method based on OCSVM algorithm using multi strategies integration

Copy number variation (CNV) is an important part of human genetic variations, which is associated with various kinds of diseases. To tackle the limitations of traditional CNV detection methods, such as restricted detection types, high error rates, and challenges in precisely identifying the location of variant breakpoints, a new method called MSCNV (copy number variations detection method for multi-strategies integration based on a one-class support vector machine model) is proposed. MSCNV establishes a multi-signal channel that integrates three strategies: read depth, split read, and read pair. First, a one-class support vector machine algorithm is used to detect abnormal signals in read depth and mapping quality values to determine the rough CNV region. Then, the rough CNV region is filtered by using paired read signals to improve the precision of MSCNV method. Finally, MSCNV explores and recognizes tandem duplication regions, interspersed duplication regions, and loss regions. It uses split read signals to determine the precise location of mutation points and to determine the type of variation. Compared with Manta, FREEC, GROM-RD, Rsicnv, and CNVkit, MSCNV significantly improves the sensitivity, precision, F1-score, and overlap density score of CNV detection while reducing the boundary bias of the detection results.

Human-in-the-loop machine learning-based quantitative assessment of hemifacial spasm based on volumetric interpolated breath-hold examination MR.

The purpose of this study was to assess the severity of hemifacial spasm (HFS) through quantitative measures that associated it with neurovascular contact (NVC). We enrolled 108 HFS patients (63 severe and 45 mild cases) and implemented a human-in-the-loop approach to develop a quantitative NVC feature package. This process involved using interactive segmentation on three-dimensional volumetric interpolated breath-hold examination (VIBE) MR images to delineate vascular and nerve structures. From these segmentations, we extracted quantitative NVC features, forming an NVC feature package, and applied a support vector machine model to assess HFS severity. Our interactive segmentation technique achieved high accuracy (Dice similarity coefficients of 0.905 ± 0.030 for vascular structures and 0.922 ± 0.086 for nerves). The NVC feature package, comprising distance between vascular structures and nerves, vascular diameter, their ratio, and clinical characteristics, enabled our model to assess HFS severity with an AUC of 0.823 (95% CI: 0.714-0.932, p < 0.001). This study introduced a quantitative approach to understanding the relationship between HFS severity and NVC, using VIBE MR imaging. Our model offers a promising tool for enhancing clinical decision-making and offers deeper insights into the impact of NVCon HFS, aiming to improve patient outcomes. Microvascular decompression is well-established as a safe and effective treatment for HFS. However, there is a gap assessing the severity of HFS using quantitative measures that directly link it to NVC. Our method introduced a quantitative and objective alternative for assessing the severity of HFS to addressing this gap.

Development and validation of a novel risk-predicted model for early sepsis-associated acute kidney injury in critically ill patients: a retrospective cohort study

ObjectivesThis study aimed to develop a prediction model for the detection of early sepsis-associated acute kidney injury (SA-AKI), which is defined as AKI diagnosed within 48 hours of a sepsis...

Multiscale 1D-CNN for Damage Severity Classification and Localization Based on Lamb Wave in Laminated Composites

Lamb-wave-based structural health monitoring is widely employed to detect and localize damage in composite plates; however, interpreting Lamb wave signals remains challenging due to their dispersive characteristics. Although convolutional neural networks (CNNs) demonstrate a significant capability for pattern recognition within these signals relative to other machine learning models, CNNs frequently encounter difficulties in capturing all the underlying patterns when the damage severity varies. To address this issue, we propose a multiscale, one-dimensional convolutional neural network (MS-1D-CNN) to assess the damage severity and localize damage in laminated plates. The MS-1D-CNN is capable of learning both low- and high-level features, enabling it to distinguish between minor and severe damage. The dataset was obtained experimentally via a sparse array of four lead zirconate titanates, with signals from twelve paths fused and downsampled before being input into the model. The efficiency of the model was evaluated using accuracy, precision, recall, and F1-score metrics for severity identification, along with the mean squared error, mean absolute error, and R2 for damage localization. The experimental results indicated that the proposed MS-1D-CNN outperformed support vector machine and artificial neural network models, achieving higher accuracy in both identifying damage severity and localizing damage with minimal error.

The predictive value of radiomics and deep learning for synchronous distant metastasis in clear cell renal cell carcinoma

ObjectiveThe objective of this research was to devise and authenticate a predictive model that employs CT radiomics and deep learning methodologies for the accurate prediction of synchronous distant metastasis (SDM) in clear cell renal cell carcinoma (ccRCC).MethodsA total of 143 ccRCC patients were included in the training cohort, and 62 ccRCC patients were included in the validation cohort. The CT images from all patients were normalized, and the tumor regions were manually segmented via ITK-SNAP software. Radiomic features were extracted via the FAE toolkit. The least absolute shrinkage and selection operator (LASSO) algorithm was employed to select features and build various machine learning models. Additionally, the largest cross-section of the tumor was cropped to train the deep learning model. Multiple deep learning models were trained to predict SDM in ccRCC patients. The results of the best machine learning model were then fused with those of the deep learning model to create a combined model.ResultsOf the 944 radiomic features identified, 15 were closely associated with SDM. With these 15 features, the support vector machine (SVM) model emerged as the most effective, demonstrating areas under the curve (AUC) of 0.860 and 0.813 in the training and validation cohort, respectively. Among the deep learning models, ResNet101 performed optimally, achieving AUC of 0.815 and 0.743 in the training and validation cohort, respectively. The combined model yielded an AUC of 0.863. Decision curve analysis suggested that the combined model offers superior clinical applicability.ConclusionThe model integrates radiomics and deep learning, showing significant potential in predicting SDM in ccRCC patients. It holds promise for supporting clinical decision-making, reducing missed diagnoses of SDM, and guiding patients in further enhancing their systemic examinations.

Prediction of implant failure risk due to periprosthetic femoral fracture after primary elective total hip arthroplasty : a simplified and validated model based on 154,519 total hip arthroplasties from the Swedish Arthroplasty Register.

While cementless fixation offers potential advantages over cemented fixation, such as a shorter operating time, concerns linger over its higher cost and increased risk of periprosthetic fractures. If the risk of fracture can be forecasted, it would aid the shared decision-making process related to cementless stems. Our study aimed to develop and validate predictive models of periprosthetic femoral fracture (PPFF) necessitating revision and reoperation after elective total hip arthroplasty (THA). We included 154,519 primary elective THAs from the Swedish Arthroplasty Register (SAR), encompassing 21 patient-, surgical-, and implant-specific features, for model derivation and validation in predicting 30-day, 60-day, 90-day, and one-year revision and reoperation due to PPFF. Model performance was tested using the area under the curve (AUC), and feature importance was identified in the best-performing algorithm. The Lasso regression excelled in predicting 30-day revisions (area under the receiver operating characteristic curve (AUC) = 0.85), while the Gradient Boosting Machine (GBM) model outperformed other models by a slight margin for all remaining endpoints (AUC range: 0.79 to 0.86). Predictive factors for revision and reoperation were identified, with patient features such as increasing age, higher American Society of Anesthesiologists grade (> III), and World Health Organization obesity classes II to III associated with elevated risks. A preoperative diagnosis of idiopathic necrosis increased revision risk. Concerning implant design, factors such as cementless femoral fixation, reverse-hybrid fixation, hip resurfacing, and small (< 35 mm) or large (> 52 mm) femoral heads increased both revision and reoperation risks. This is the first study to develop machine-learning models to forecast the risk of PPFF necessitating secondary surgery. Future studies are required to externally validate our algorithm and assess its applicability in clinical practice.

Predicting Polymerase Chain Reaction Success: Integrating the K-Word Order Model, Physicochemical Properties Modeling of Double Bases, and Support Vector Machine.

Polymerase Chain Reaction (PCR) has been a pivotal scientific technique since the twentieth century, and it is widely applied across various domains. Despite its ubiquity, challenges persist in efficiently amplifying specific DNA templates. While PCR experimental procedures have garnered significant attention, the analysis of the DNA template, which is the experiment's focal point, has been notably overlooked. This study addresses the uncertainty surrounding the amplification of DNA fragments using conventional Taq DNA polymerase-based PCR protocols. The imperative need to characterize DNA templates and devise a reliable method for predicting PCR success is underscored. In this study, we formulate a 72-dimensional feature vector representing a DNA template through the utilization of k-word order and modeling of physicochemical properties of double bases. Subsequently, a Support Vector Machine (SVM) model is employed to assess PCR results. A jackknife cross-validation test is used to evaluate the anticipated success rates, resulting in an overall accuracy of 95.77%. Sensitivity, specificity, and Matthew's Correlation Coefficient (MCC) stand at 95.75%, 95.79%, and 0.915, respectively.

Rice Extent and Cropping Patterns in Terengganu Malaysia Based on Sentinel-2 Data on Google Earth Engine

Rice is a vital staple food in Malaysia, with consumption of 2.7 million MT in 2016, forecasted to rise to 3.5 million MT by 2026. To address food security, the Malaysian government targets a 70% self-sufficiency level (SSL) in rice production, requiring precise spatiotemporal information on rice cultivation. Remote sensing has been widely used to map rice extent in Malaysia, particularly in granary areas (facilitated by irrigation system), the main production zones. However, studies on non-granary areas (without irrigation systems) remain limited. This study addresses this gap by employing the Phenology-Expert Based Unsupervised Classification Method (PEB-UC) with Sentinel-2 time series data on the Google Earth Engine platform to map rice extent and cropping patterns at the sub-district level across Terengganu State, Malaysia, covering both granary and non-granary areas. The results revealed scattered rice parcels totalling 8,184 ha, with 4,377 ha in the IADA KETARA granary area (Besut District) and 3,807 ha in non-granary areas. The maps showed a relative discrepancy of -29.15% with agricultural statistics but demonstrated a strong correlation at the district level (R2 = 0.99; RMSE = 632 ha). Validation of PEB-UC achieved an overall accuracy of 0.979 and a kappa coefficient of 0.957, outperforming Random Forest (RF) and Support Vector Machine (SVM) models. The PEB-UC rice map displayed denser, clearer separability between rice and non-rice compared to supervised models, as shown in the comparison map at https://rudiyanto.users.earthengine.app/view/riceterengganu. This study provides valuable insights into rice cultivation in Terengganu State and supports efforts to enhance food security.

Machine learning analysis of glutamate receptor activity in developing locus coeruleus neurons

Purpose/aim: The developing brain undergoes a remarkable process of synaptic changes. Material and methods: To investigate the developmental changes in glutamatergic synaptic connections using the whole-cell patch clamp method, evoked excitatory postsynaptic currents (eEPSCs) were recorded from locus coeruleus (LC) neurons, a brain region crucial for cognitive functions, in rats at ages 7, 14, and 21 days. We employed fractal analysis to compute fractal dimensions of AMPA and NMDA receptors, serving as markers for synaptic maturation. Results: Our findings revealed a significant increase in fractal dimensions during the third postnatal week and hence a developmental chenge of synaptic connections. A strong positive correlation between amplitude and fractal dimensions, in Pearson correlation analysis suggested that the synaptic currents’ amplitude is closely related to the fractal properties of the receptors. A linear relationship between fractal dimensions and age indicated that fractal analysis can be a robust tool for predicting developmental changes. Additionally, we employed machine learning techniques to predict developmental changes based on AMPA and NMDA receptors. Support Vector Machine (SVM) models outperformed random forest models in accurately predicting age-dependent developmental changes, as indicated by the area under the curve (AUC) values. SVM achieved an AUC of 0.89 for AMPA receptors and 0.86 for NMDA receptors, demonstrating the effectiveness of fractal-based features in characterizing synaptic maturation. Conclusion: This study offers valuable insights into synaptic development in the LC nucleus and demonstrates the potential of fractal analysis as a tool to understand brain plasticity and early development. Fractal dimensions play a crucial role in characterizing the maturation of glutamatergic synapses and neural circuitry development.

Research on Coal and Gas Outburst Prediction and Sensitivity Analysis Based on an Interpretable Ali Baba and the Forty Thieves–Transformer–Support Vector Machine Model

Coal and gas outbursts pose significant threats to underground personnel, making the development of accurate prediction models crucial for reducing casualties. By addressing the challenges of highly nonlinear relationships among predictive parameters, poor interpretability of models, and limited sample data in existing studies, this paper proposes an interpretable Ali Baba and the Forty Thieves–Transformer–Support Vector Machine (AFT-Transformer-SVM) model with high predictive accuracy. The Ali Baba and the Forty Thieves (AFT) algorithm is employed to optimise a Transformer-based feature extraction, thereby reducing the degree of nonlinearity among sample data. A Transformer-SVM model is constructed, wherein the Support Vector Machine (SVM) model provides negative feedback to refine the Transformer feature extraction, enhancing the prediction accuracy of coal and gas outbursts. Various classification assessment methods, such as TP, TN, FP, FN tables, and SHAP analysis, are utilised to improve the interpretability of the model. Additionally, the permutation feature importance (PFI) method is applied to conduct a sensitivity analysis, elucidating the relationship between the sample data and outburst risks. Through a comparative analysis with algorithms such as eXtreme gradient boosting (XGBoost), k-nearest neighbour (KNN), radial basis function networks (RBFNs), and Bayesian classifiers, the proposed method demonstrates superior accuracy and effectively predicts coal and gas outburst risks, achieving 100% accuracy in the sample dataset. The influence of parameters on the model is analysed, highlighting that the coal seam gas content is the primary factor driving the outburst risks. The proposed approach provides technical support for coal and gas outburst predictions across different mines, enhancing emergency response and prevention capabilities for underground mining operations.