Lower Mean Square Error Research Articles

Optimizing critical quality attributes of fast disintegrating tablets using artificial neural networks: a scientific benchmark study

Objective The objective of this study is to create predictive models utilizing machine learning algorithms, including Artificial Neural Networks (ANN), k-nearest neighbor (kNN), support vector machines (SVM), and linear regression, to predict critical quality attributes (CQAs) such as hardness, friability, and disintegration time of fast disintegrating tablets (FDTs). Methods A dataset of 864 batches of FDTs was generated by varying binder types and amounts, disintegrants, diluents, punch sizes, and compression forces. Preprocessing steps included normalizing numerical features based on industry standards, one-hot encoding for categorical variables, and addressing outliers to ensure data consistency. Four machine learning models were trained and evaluated on R2 values and mean squared error (MSE). Feature importance was analyzed using permutation importance, and statistical validation (p < 0.05) and confidence intervals were computed for model performance. The ‘differential_evolution’ function was used to optimize the formulation. Results Among the models, ANN demonstrated the highest predictive accuracy, achieving R2 values up to 0.9550 with the lowest MSE across training and test datasets, outperforming kNN, SVM, and linear regression. The ANN’s ability to model complex, non-linear interactions between formulation variables was statistically significant, as validated through six checkpoint batches of acetylsalicylic acid FDTs. The feature importance analysis revealed compression force, binder type, and punch size as the most influential factors affecting hardness, while disintegrant type influenced friability. The ‘differential_evolution’ function effectively optimized the CQAs, resulting in FDTs with ideal characteristics. Conclusion The ANN model, integrated with differential evolution, provided a robust tool for optimizing FDT formulations by accurately predicting CQAs and reducing the need for extensive experimental trials. Compared to traditional optimization methods, ANN excels in capturing intricate multi-variable relationships, making it a valuable approach for scaling beyond acetylsalicylic acid to other formulations. This method enhances the consistency and efficiency of tablet formulation, supporting broader pharmaceutical applications

Iterative frequency-domain equalization with multipath interference cancellation for aeronautical telemetry

The increasing transmitted signal bandwidth of aeronautical telemetry systems leads to the channel’s frequency selectivity. Equalization techniques are always adopted to conquer such problems. To optimize the equalization performance, we pursue minimizing the mean square error (MSE) of the equalizers. The solution shows that equalizing the line-of-sight (LOS) signal results in a lower MSE than equalizing the received signal directly. Thus, we proposed an iterative equalization algorithm with multipath interference cancellation (MIC) in this letter. In each iteration, the MIC is used to extract the LOS signal from the received signal. Then, we equalize the LOS signal to reduce the MSE of the equalization. Finally, the symbols are detected from the equalized LOS signal to achieve bit-error-rate (BER) performance improvement. The numerical results demonstrate that the proposed method provides a superior BER performance with lower complexity compared to the state-of-the-art equalizers.

Enhancing Estimation Accuracy of Prostate Cancer VMAT Planning: A Knowledge-based Approach Using Multiple Collimator Angles.

This study aimed to determine whether a knowledge-based planning model incorporating treatment plans with multiple collimator angles (Multi-coll. model) provides superior estimation accuracy and plan quality for a range of collimator angles compared with a typical institutional model (Inst. model). Five institutions using volumetric modulated arc therapy for prostate cancer participated in the study. The Inst. model comprised plans using a fixed collimator angle, whereas the Multi-coll. model registered plans using single arcs and collimator angles of 10°, 30°, 50°, and 70°. The coefficient of determination and mean squared error were calculated, and the estimation accuracy and dosimetric results for three validation cases at each institution were compared for each model. Dosimetric parameters included volumes receiving at least 20%, 50%, and 90% of the prescription dose (V20, V50, and V90, respectively) at each collimator angle. The Multi-coll. model yielded a higher coefficient of determination and lower mean squared error than the Inst. model, although overfitting of data was a concern at two institutions. The mean absolute errors between estimated and calculated values were (Inst. model vs. Multi-coll. model): 4.41% vs. 2.60% (V20), 4.03% vs. 2.27% (V50), and 2.07% vs. 1.09% (V90) for the rectum; 3.48% vs. 3.42% (V20), 2.24% vs. 3.65% (V50), and 1.11% vs. 0.73% (V90) for the bladder. The Multi-coll. model exhibited a lower rectal and vesical V20 than the Inst. The Multi-coll. model had a greater estimation accuracy and slightly higher plan quality than the Inst.

Data-Driven Models for Significant Wave Height Forecasting: Comparative Analysis of Machine Learning Techniques

Accurate prediction of significant wave height (SWH) is critical for coastal safety, marine operations, and disaster management. Traditional numerical models for wave prediction are computationally intensive and often lack accuracy, prompting a shift towards data-driven methods. This study explores the efficacy of machine learning (ML) models in forecasting SWH in the coastline of North Stradbroke Island, Queensland, Australia. Using a dataset spanning 2010 to 2022, the study employs wave characteristics such as maximum wave height (Hmax), wave periods (Tz, Tp), peak direction, and sea surface temperature (SST) as predictors. Three ML models—Linear Regression, Decision Tree, and Random Forest—were trained and evaluated using performance metrics like Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-squared (R²). The Random Forest model demonstrated the best predictive accuracy with the lowest MSE (0.0074) and highest R² (0.958), outperforming both Linear Regression and Decision Tree models. This improvement in prediction accuracy supports the model's application for coastal management, ensuring better forecasting of wave conditions. The proposed approach shows significant advantages, including better handling of non-linearities and reduced computational costs compared to conventional numerical methods, such as SWAN and WAM, making it highly applicable for real-time wave forecasting, particularly for regions with complex coastal dynamics such as North Stradbroke Island.

Unlimited sampling beyond modulo

Analog-to-digital converters (ADCs) act as a bridge between the analog and digital domains. Two important attributes of any ADC are sampling rate and its dynamic range. For bandlimited signals, the sampling should be above the Nyquist rate. It is also desired that the signals' dynamic range should be within that of the ADC's; otherwise, the signal will be clipped. Nonlinear operators such as modulo or companding can be used prior to sampling to avoid clipping. To recover the true signal from the samples of the nonlinear operator, either high sampling rates are required, or strict constraints on the nonlinear operations are imposed, both of which are not desirable in practice. In this paper, we propose a generalized flexible nonlinear operator which is sampling efficient. Moreover, by carefully choosing its parameters, clipping, modulo, and companding can be seen as special cases of it. We show that bandlimited signals are uniquely identified from the nonlinear samples of the proposed operator when sampled above the Nyquist rate. Furthermore, we propose a robust algorithm to recover the true signal from the nonlinear samples. Compared to the existing methods, our approach has a lower mean-squared error for a given sampling rate, noise level, and dynamic range. Our results lead to less constrained hardware design to address the dynamic range issues while operating at the lowest rate possible.

Predictive Modeling for Enhancing MOOC Completion Rates: A Case Study

In the realm of online asynchronous learning platforms, accurately tracking student performance to predictcourse completion times poses a significant challenge. Completion rates for MOOCs are typically low, with abias towards participants with higher education levels. Understanding factors such as student motivation,engagement, participation, and learning pathway design is crucial for improving student outcomes in onlinecourses. This research developed a predictive framework utilizing advanced deep learning techniques toaccurately forecast course completion times for participants enrolled in an introductory programming course(&amp;quot;Python for Beginners&amp;quot; course on the Open Learning Platform of University of Moratuwa Sri Lanka). Byaccurately tracking student performance and leveraging a diverse dataset encompassing demographic andeducational variables, the research seeks to identify factors influencing course completion and predictindividual student outcomes. By utilising deep learning techniques, the prediction performance of the modelwill be improved, ultimately contributing to a more precise forecast of course completion times forparticipants. Evaluation of the model resulted in low Mean Absolute Error (MAE) of 0.0080 and low MeanSquared Error (MSE) of 0.0033 which promises the effectiveness of the developed method in accuratelypredicting course completion times for students. The findings of this study may help increase the successfulcompletion rate of such courses which are delivered in the online asynchronous mode. The study employedadvanced deep learning models optimized through Bayesian methods, highlighting the potential of thesetechniques to enhance MOOC completion rates by offering precise forecasts and actionable insights intostudent engagement. The comprehensive analysis revealed that variables such as &amp;#39;Current_Lesson&amp;#39;, &amp;#39;SessionTime Category&amp;#39;, and &amp;#39;District_Score&amp;#39; significantly influence completion times. The robust methodologicalframework, including feature engineering, model training, and hyperparameter optimization, sets a precedentfor future research in the field. This research contributes to educational data mining and predictive analytics,offering a scalable approach to improving completion rates and educational outcomes across various onlinelearning platforms. Future research should explore incorporating real-time data and longitudinal studies toenhance model accuracy and generalizability. Additionally, addressing potential biases in the dataset, such asdemographic, prior knowledge, and resource access disparities, is essential to ensure the fair and equitableapplication of the model across diverse student populations. Expanding the research to include a wider rangeof courses and institutions will further validate the model&amp;#39;s robustness and applicability in differenteducational contexts.

Incorporating driving behavior into vehicle fuel consumption prediction: methodology development and testing

This study is a significant endeavor involving the development and testing of a comprehensive methodology to incorporate driving behavior into the analysis and prediction of vehicle fuel consumption. It underscores the crucial importance of understanding how different driving behavior affect fuel efficiency. The framework we present is a theoretical construct and a practical tool. It provides a robust, multi-step process for linking driving behavior to fuel consumption, leveraging both traditional statistical methods and advanced machine learning techniques to derive actionable insights. To test the framework, we used a naturalistic data that includes about 5408 different road users in a mixed traffic environment and urban settings in Germany. We applied a microscopic fuel consumption model to calibrate the framework and an unsupervised clustering algorithm to classify the behavior of the driver interacting with each other and with vulnerable road users. The framework includes developing Linear regression model as a baseline, which yields an R-squared of 0.511 and a Mean Squared Error (MSE) of 0.031, indicating moderate predictive accuracy. The final step includes choosing Random Forest as a better model, which achieves a higher R-squared of 0.956 and a lower MSE of 0.003. We also found that conservative and aggressive driving leads to significantly higher and more discrepancy in fuel consumption than normal driving behavior. These insights can promote more efficient driving practices, leading to significant fuel savings and environmental benefits.

Transient pressure prediction in large-scale underground natural gas storage: A deep learning approach and case study

Large-scale underground natural gas storage (UNGS) facilities typically allocate only one month annually for well shut-in to facilitate pressure build-up. This limited well shut-in period affects the subsequent pressure measurements and evaluation of gas storage capacity. This study proposes a novel workflow to directly predict the transient pressure behavior of UNGS during pressure build-up without requiring additional well shut-in operations. The proposed workflow uses the gas injection-withdrawal rate as a dynamic input feature for a deep learning model, with reservoir pressure as the output feature. An enhanced WA-BiLSTM model integrates multiple physical mechanisms and advanced optimization algorithms. The model achieves mean squared error (MSE) of 2 × 10-3, which is less than 5 % of traditional model's results. Field data from the largest Hutubi UNGS exhibit a prediction accuracy of 99.0 % during the well shut-in phase. High-precision prediction results with low MSE ensure the reliability of pressure derivative data. By analyzing the predicted pressure data, the actual gas storage volume is 104.75 × 108 m3, accounting for 97.9 % of the facility's designated storage capacity.

Removal of zinc from wastewaters using Turkish bentonite and artificial neural network [ANN] modeling

In this study, Ordu-Unye bentonite was used as an adsorbent in the removal of zinc from aqueous solutions. The aim of the experimental part of the study was to ascertain how zinc removal was affected by variables such as pH, adsorbent amount, contact time, and initial zinc concentration. In the second part of the experiments, bentonite was modified with two different acids and the adsorption performance of modified bentonite was also investigated. Characterization of raw and modified bentonites was also carried out using FTIR and XRD. It was observed that acid modification of bentonite negatively affected the zinc removal process from aqueous solutions. In this study, higher zinc removal (95 %) was obtained with raw bentonite compared to acid modified bentonites (58.4 % in HNO3 activated, 43.8 % for H2SO4 activated). Equilibrium isotherms were obtained and modelled to explain the adsorption mechanism. Adsorption isotherm studies showed that zinc adsorption fits well with Langmuir (R2: 0.99) and Temkin (R2: 0.97) models. Besides from these experimental investigations, various artificial neural network (ANN) training techniques were used to optimize the zinc adsorption process. By trial and error, the optimal performance was obtained by changing the number of hidden neurons in each layer of the neural network architecture. These models under study were analyzed to determine their R2 and mean square error (MSE) values, and the optimal outcomes were identified. Among the various training models of ANN, it was determined that the Bayesian Regularization method exhibited the optimum network architecture with the highest R2 (R2:0.995) and lowest MSE (MSE:0.0008) ratio.

An efficient water quality index forecasting and categorization using optimized Deep Capsule Crystal Edge Graph neural network.

The world's freshwater supply, predominantly sourced from rivers, faces significant contamination from various economic activities, confirming that the quality of river water is critical for public health, environmental sustainability, and effective pollution control. This research addresses the urgent need for accurate and reliable water quality monitoring by introducing a novel method for estimating the water quality index (WQI). The proposed approach combines cutting-edge optimization techniques with Deep Capsule Crystal Edge Graph neural networks, marking a significant advancement in the field. The innovation lies in the integration of a Hybrid Crested Porcupine Genghis Khan Shark Optimization Algorithm for precise feature selection, ensuring that the most relevant indicators of water quality (WQ) are utilized. Furthermore, the use of the Greylag Goose Optimization Algorithm to fine-tune the neural network's weight parameters enhances the model's predictive accuracy. This dual optimization framework significantly improves WQI prediction, achieving a remarkable mean squared error (MSE) of 6.7 and an accuracy of 99%. By providing a robust and highly accurate method for WQ assessment, this research offers a powerful tool for environmental authorities to proactively manage river WQ, prevent pollution, and evaluate the success of restoration efforts. PRACTITIONER POINTS: Novel method combines optimization and Deep Capsule Crystal Edge Graph for WQI estimation. Preprocessing includes data cleanup and feature selection using advanced algorithms. Deep Capsule Crystal Edge Graph neural network predicts WQI with high accuracy. Greylag Goose Optimization fine-tunes network parameters for precise forecasts. Proposed method achieves low MSE of 6.7 and high accuracy of 99%.

Integrating machine learning and deep learning for enhanced supplier risk prediction

The importance of anticipating and preventing disruptions is underscored by the increased operational complexity and vulnerability caused by advancements in supply chain management (SCM). This has spurred interest in integrating machine learning (ML) and deep learning (DL) into supply chain risk management (SCRM). In this paper, we introduce a tailored method using ML and DL to improve SCRM by predicting supplier failures, thus boosting efficiency and resilience in SC operations. Our method involves five phases focused on classifying and predicting supplier failures in non-conforming deliveries. This involves forecasting failure quantities and estimating total disruption costs. Initially, data from an automotive company is selected, and appropriate potential features and algorithms are selected, performance metric aligns with case study objectives, facilitating method evaluation are used such as: Precision, recall, F1-score, and accuracy metrics assess classification models, while Mean Squared Error (MSE) is used for regression tasks. Finally, an experimental design optimizes models, assessing success rates of various algorithms and their parameters within the chosen feature space. Experimental results underscore the success of our methodology in model development. In the classification task, the Random Forest (RF) classifier achieved 86% accuracy. When combined with the Gradient Boosting classifier, the ensemble exhibited enhanced accuracy, highlighting the complementary strengths of both algorithms and their synergistic impact, surpassing the performance of RF, Support Vector Regression (SVR), k-Nearest Neighbors (KNN), and Artificial Neural Network (ANN). Noteworthy is the performance in regression tasks, where Linear Regression, ANN, and RF Regressor displayed exceptionally low MSE compared to other models.

Machine Learning Algorithms for Prediction of Boiler Steam Production

The continuous increase in global electricity demand has resulted in boiler power plants becoming a significant energy source. The production of steam is a principal indicator of boiler efficiency, and the accurate prediction of steam production is paramount importance for the enhancement of boiler efficiency and the reduction of operational costs. In this study employs a boiler dataset with a steam production capacity of 420 tons per hour. A total of 25 independent variables were extracted from the original 39 variables through data processing and feature engineering for the purpose of prediction analysis. Subsequently, 8 machine learning models were used for modeling predictions. Grid search cross-validation was employed in order to optimise the performance of the model. The models were analysed and assessed using the Mean Squared Error (MSE) metrics. The results show that random forest achieves the highest accuracy among the 8 single models. Based on 8 models, New Bagging ensemble model is proposed, which combined predictions from 8 single models, demonstrated the optimal overall fit and the lowest MSE, achieved the purpose of the research. The present study demonstrates the ability to analyse and predict complex industrial systems with machine learning algorithms, and provides insights into the use of machine learning algorithms for industrial big data analytics and Industry 4.0. Further work could explore using larger datasets and deep learning to make predictions more accurate.

Improve Metode Lightgbm untuk Prediksi Harga Mobil Bekas Menggunakan Hyper-Parameter Tuning

This study aims to predict used car prices using the LightGBM method and hyperparameter tuning techniques in the context of data science. The analysis process includes collecting historical data on used cars, preprocessing the data to clean and encode variables, and splitting the data into training and testing sets. The LightGBM model was trained and optimized through hyperparameter tuning using GridSearchCV to improve model performance. The model was evaluated using metrics such as Mean Squared Error (MSE) and R-squared. The results indicate that the well-optimized LightGBM model can accurately predict used car prices with high accuracy. The low MSE value (35207938112.028404) and high R-squared value (0.9462871489515565) demonstrate the model's excellent predictive quality. This research provides deeper insights into the factors influencing used car prices and contributes to the development of effective and reliable predictive models.

Comparative Analysis of Linear Regression, Decision Tree, and Gradient Boosting Models for Predicting Drug Corrosion Inhibition Efficiency Using QSAR Descriptors

&lt;table width="593" border="1" cellspacing="0" cellpadding="0"&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td valign="top" width="388"&gt;&lt;p&gt;Corrosion in industrial environments poses significant economic and safety challenges, necessitating the development of effective inhibitors. Organic compounds, particularly pharmaceuticals, have emerged as promising corrosion inhibitors due to their efficiency and environmental benefits. However, predicting these compounds' corrosion inhibition efficiency (CIE) remains complex and requires advanced computational methods. This study investigates the predictive capabilities of three machine learning (ML) models, namely linear regression, decision tree, and gradient boosting regression, using Quantitative Structure-Activity Relationship (QSAR) descriptors. A dataset containing 14 QSAR descriptors was compiled from experimental studies on various pharmaceutical-based inhibitors. The dataset was divided into training (90%) and testing (10%) subsets to evaluate model performance. The research follows the CRISP-DM methodology, a systematic framework that includes data preparation, model training, and evaluation. Key performance metrics, including Mean Squared Error (MSE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and the coefficient of determination (R²), were used to assess model accuracy. Among the models, Gradient Boosting Regression achieved the most promising results, with the lowest MSE (21.52) and the highest R² (0.21), reflecting its ability to capture non-linear relationships in the data. Despite the relatively modest R², this model demonstrates the potential for improving computational approaches to corrosion inhibition prediction. This study highlights the value of machine learning in optimizing the selection of corrosion inhibitors, potentially reducing the reliance on extensive laboratory testing and accelerating the discovery of efficient, eco-friendly solutions for industrial applications.&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt;

Ternary blended concrete strength evaluation: experimental and artificial intelligence techniques

Purpose The purpose of this study is to forecast the mechanical properties of ternary blended concrete (TBC) modified with oyster shell powder (OSP) and shea nutshell ash (SNA) using deep neural network (DNN) models. Design/methodology/approach DNN models with three hidden layers, each layer containing 5–30 nodes, were used to predict the target variables (compressive strength [CS], flexural strength [FS] and split tensile strength [STS]) for the eight input variables of concrete classes 25 and 30 MPa. The concrete samples were cured for 3–120 days. Levenberg−Marquardt's backpropagation learning technique trained the networks, and the model's precision was confirmed using the experimental data set. Findings The DNN model with a 25-node structure yielded a strong relation for training, validating and testing the input and output variables with the lowest mean squared error (MSE) and the highest correlation coefficient (R) values of 0.0099 and 99.91% for CS and 0.010 and 98.42% for FS compared to other architectures. However, the DNN model with a 20-node architecture yielded a strong correlation for STS, with the lowest MSE and the highest R values of 0.013 and 97.26%. Strong relationships were found between the developed models and raw experimental data sets, with R2 values of 99.58%, 97.85% and 97.58% for CS, FS and STS, respectively. Originality/value To the best of the authors’ knowledge, this novel research establishes the prospects of replacing SNA and OSP with Portland limestone cement (PLC) to produce TBC. In addition, predicting the CS, FS and STS of TBC modified with OSP and SNA using DNN models is original, optimizing the time, cost and quality of concrete.

Comparative study of five machine learning algorithms on prediction of the height of the water-conducting fractured zone in undersea mining

Prediction of water-conducting fractured zone (WCFZ) of mine overburden is the premise for reducing or eliminating water inrush hazards in undersea mining. To obtain a more robust and precise prediction of WCFZ in undersea mining, a WCFZ prediction dataset with 122 cases of fractured zones was constructed. Five machine learning algorithms (linear regression, XGBRegressor, RandomForestRegressor, LineareSVR, and KNeighborsRegressor) were employed to develop five corresponding predictive models, taking multiple factors into account.The optimal parameters for each model are obtained through ten-fold cross-validation (10CV). The model's predictive performance was validated and assessed using two metrics, namely the coefficient of determination (R2) and mean squared error (MSE). A comparison was made with the regression performance of commonly used empirical formulas. The results indicate that the constructed model outperforms reliance solely on theoretical criteria, showing a high R2 value of up to 0.925 and a low MSE value of 3.61. The proposed model was validated in a recently established mining area on Sanshan Island, China. It shows low absolute and relative errors of 0.71 m and 2.01%, respectively, between the predicted value from the model and observation result from the field, demonstrating a high level of consistency with on-site conditions. This paves a path to leveraging machine learning algorithms for predicting the height of WCFZ.

Artificial Intelligence in Advance Angiogenesis and Inflammation Research: A Breakthrough in Disease Prediction and Therapy

Background: Angiogenesis and inflammation are fundamental biological processes, crucial to human health. Dysregulation of these processes is implicated in diseases such as cancer, cardiovascular disorders, and autoimmune diseases. Despite advances in research, the complexity of these interactions has been challenging to understand. Recent developments in artificial intelligence (AI) offer a promising approach for overcoming these challenges, especially in big data analysis. This study explores AI applications in quantifying angiogenesis and inflammatory markers and predicting disease progression. Methods: AI algorithms, including machine learning (ML) and deep learning (DL), were employed to analyze high-throughput biological data. The study applied Lasso regression for biomarker discovery, Long Short-Term Memory (LSTM) networks for predicting disease progression, and Gaussian Mixture Models (GMM) for patient subgroup identification. Image analysis using DeepLabv3+ was conducted to assess angiogenesis and inflammatory markers in histological images. Model performance was evaluated using R-squared (R²), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) metrics. Results: The AI framework demonstrated high accuracy in predicting disease progression, with notable R² values and low MSE and RMSE values. The application of AI led to the successful identification of angiogenesis-related genes and biomarkers in various diseases, including diabetic foot ulcers and chronic obstructive pulmonary disease. AI-based image analysis also provided precise quantification of angiogenesis and inflammation, enhancing the understanding of disease mechanisms. Conclusion: AI-driven approaches significantly improve the analysis of complex biological processes, offering new insights into angiogenesis and inflammation. The high predictive accuracy of the AI models underscores their potential in clinical applications, such as personalized treatment strategies and disease monitoring. As AI continues to evolve, its integration into biomedical research will likely yield further advancements in disease prediction, diagnosis, and treatment.

Group Inverse-Gamma Gamma Shrinkage for Sparse Linear Models with Block-Correlated Regressors

Heavy-tailed continuous shrinkage priors, such as the horseshoe prior, are widely used for sparse estimation problems. However, there is limited work extending these priors to predictors with grouping structures. Of particular interest in this article, is regression coefficient estimation where pockets of high collinearity in the covariate space are contained within known covariate groupings. To assuage variance inflation due to multicollinearity we propose the group inverse-gamma gamma (GIGG) prior, a heavy-tailed prior that can trade-off between local and group shrinkage in a data adaptive fashion. A special case of the GIGG prior is the group horseshoe prior, whose shrinkage profile is correlated within-group such that the regression coefficients marginally have exact horseshoe regularization. We show posterior consistency for regression coefficients in linear regression models and posterior concentration results for mean parameters in sparse normal means models. The full conditional distributions corresponding to GIGG regression can be derived in closed form, leading to straightforward posterior computation. We show that GIGG regression results in low mean-squared error across a wide range of correlation structures and within-group signal densities via simulation. We apply GIGG regression to data from the National Health and Nutrition Examination Survey for associating environmental exposures with liver functionality.

Multi-task metric learning for optical performance monitoring

In our experiments, applying few shot metric learning for optical performance monitoring (OPM), we set the dataset as 16-way-6-shot. Modulation format identification (MFI) was utilized as a classification task, and optical signal-to-noise ratio (OSNR) estimation was used as a regression task for joint analysis. Multi-task metric learning (MML) used the adaptive weights to balance the weights of the three metric functions, six modulation formats (QPSK, 8QAM, 16QAM, 32QAM, 64QAM, 128QAM) are classified correctly with 100 % accuracy after 200 epochs. Furthermore, the lowest mean square error (MSE) of OSNR is 0.431 dB. Then, Ablation experiments compute the corresponding similarity (SIM) for each metric function show that the MSE of MML, SIMLocal+Cosine, SIMCosine+Point, SIMLocal+Point, single-task metric learning (SML) and adaptive multi-task learning (AMTL) is 0.431 dB, 0.572 dB, 0.569 dB, 0.567 dB, 0.637 dB, 1.319 dB, respectively. The proposed model achieves the highest accuracy in MFI and the lowest MSE in OSNR estimation. Finally, when comparing the various metric functions while altering the transmission distance of the optical fiber, it was observed that MML stayed within an acceptable range between 200 km and 800 km. This shows that our algorithm requires only a small number of training samples to create a reasonably good model, offering a new approach to solving problems that arise in optical performance monitoring.

Weighted penalized m-estimators in robust ridge regression: an application to gasoline consumption data

ABSTRACT The OLS and ridge regression (RR) estimators are adversely affected, when the problem of multicollinearity and y-direction outliers occur together. The robust ridge regression with penalized parameters offers biased estimators with lower variance than OLS and RR. The optimal penalized parameter value is crucial for balancing variance and bias. This study proposes three new weighted robust penalized M-estimators to address both multicollinearity and y-direction outliers. Their performance is evaluated against OLS, RR and existing penalized M-estimators using Monte Carlo simulations based on mean squared error (MSE). The new estimators exhibit lower MSE in the presence of multicollinearity and y-direction outliers. A real data application on gasoline consumption demonstrates the superior performance of the new estimators.