Best Model In Terms Research Articles

A Bayesian spatially explicit estimation of daily egg production: application to anchovy in the Bay of Biscay

Biomass estimates of fish resources by the daily egg production method (DEPM) are sensitive to the high variability of the daily egg production ( P0) and egg mortality (Z) in space. This work presents a Bayesian approach to estimate these parameters. A prior distribution of Z based on literature serves to overcome the biologically implausible Z estimates that can result from frequentist approaches. In addition to the classical estimation of a single P0 over the spawning area, the Bayesian framework allows also the modelling of egg densities in space, by including either spatial random effects, smoothing functions, or kriging like models, providing insights into the spatial variability of P0. The Bayesian approach was applied to the Bay of Biscay anchovy DEPM surveys. Results showed that this Bayesian approximation solved the implausible Z problem resulting in tighter credible intervals of both P0 and Z. Overall, spatial models outperformed the non-spatial model in terms of goodness of fit and resulted in slightly different total production estimates across models for each year, with a moderate decrease on uncertainty estimates.

Classifying Participant Standing and Sitting Postures Using Channel State Information

Recently, channel state information (CSI) has been identified as beneficial in a wide range of applications, ranging from human activity recognition (HAR) to patient monitoring. However, these focused studies have resulted in data that are limited in scope. In this paper, we investigate the use of CSI data obtained from an ESP32 microcontroller to identify participants from sitting and standing postures in a many-to-one classification. The test is carried out in a controlled isolated environment to establish whether a pre-trained model can distinguish between participants. A total of 15 participants were recruited and asked to sit and stand between the transmitter (Tx) and the receiver (Rx), while their CSI data were recorded. Various pre-processing algorithms and techniques have been incorporated and tested on different classification algorithms, which have gone through parameter selection to enable a consistent testing template. Performance metrics such as the confusion matrix, accuracy, and elapsed time were captured. After extensive evaluation and testing of different classification models, it has been established that the hybrid LSTM-1DCNN model has an average accuracy of 84.29% and 74.13% for sitting and standing postures, respectively, in our dataset. The models were compared with the BedroomPi dataset and it was found that LSTM-1DCNN was the best model in terms of performance. It is also the most efficient model with respect to the time elapsed to sit and stand.

Assessment of Computational Fluid Dynamics Reynolds-Averaged Navier–Stokes Models for Bluff Bodies Aerodynamics

<div>Since the steady-state computational fluid dynamics (CFD) Reynolds-averaged Navier–Stokes (RANS) turbulence models offer low-cost and sensible accuracy, they are frequently utilized for bluff bodies’ external aerodynamics investigations (e.g., upwind, crosswind, and shape optimization). However, no firm certainty is made regarding the best model in terms of accuracy and cost. Based on cost and accuracy aspects, four RANS turbulence models were studied, which are Spalart–Allmaras, realizable k-ε, RNG k-ε, and SST k-ω. Ahmed body with a 25° slant angle benchmark case was introduced for this investigation. Two grids were generated to satisfy the near-wall treatment of each turbulence model. All grid settings were proposed and discussed in detail. Fluid-structure analysis was performed on five different planes. Regarding flow field prediction, realizable k-ε and renormalization group (RNG) k-ε models demonstrated a remarkable consistency with experimental data, while Menter’s shear stress transport (SST) k-ω showed a poor agreement. The obtained computational values of drag and lift coefficients were compared with experimental results. All investigated RANS turbulence models had reported results in excellent agreement with experimental drag coefficient values. The SST k-ω model has underestimated lift coefficient value with an error of about −45% with experimental value. Only realizable k-ε and RNG k-ε presented an error <10% for predicting drag and lift coefficients.</div>

DDD TinyML: A TinyML-Based Driver Drowsiness Detection Model Using Deep Learning.

Driver drowsiness is one of the main causes of traffic accidents today. In recent years, driver drowsiness detection has suffered from issues integrating deep learning (DL) with Internet-of-things (IoT) devices due to the limited resources of IoT devices, which pose a challenge to fulfilling DL models that demand large storage and computation. Thus, there are challenges to meeting the requirements of real-time driver drowsiness detection applications that need short latency and lightweight computation. To this end, we applied Tiny Machine Learning (TinyML) to a driver drowsiness detection case study. In this paper, we first present an overview of TinyML. After conducting some preliminary experiments, we proposed five lightweight DL models that can be deployed on a microcontroller. We applied three DL models: SqueezeNet, AlexNet, and CNN. In addition, we adopted two pretrained models (MobileNet-V2 and MobileNet-V3) to find the best model in terms of size and accuracy results. After that, we applied the optimization methods to DL models using quantization. Three quantization methods were applied: quantization-aware training (QAT), full-integer quantization (FIQ), and dynamic range quantization (DRQ). The obtained results in terms of the model size show that the CNN model achieved the smallest size of 0.05 MB using the DRQ method, followed by SqueezeNet, AlexNet MobileNet-V3, and MobileNet-V2, with 0.141 MB, 0.58 MB, 1.16 MB, and 1.55 MB, respectively. The result after applying the optimization method was 0.9964 accuracy using DRQ in the MobileNet-V2 model, which outperformed the other models, followed by the SqueezeNet and AlexNet models, with 0.9951 and 0.9924 accuracies, respectively, using DRQ.

Semi-Supervised Semantic Role Labeling with Bidirectional Language Models

The recent success of neural networks in NLP applications has provided a strong impetus to develop supervised models for semantic role labeling (SRL) that forego the requirement for extensive feature engineering. Recent state-of-the-art approaches require high-quality annotated datasets that are costly to obtain and almost unavailable for low-resource languages. We present a semi-supervised approach that utilizes both labeled and unlabeled data to provide performance improvement over a mere supervised SRL model. We show that our proposed semi-supervised SRL model provides larger improvement over a supervised model in the scenario where labeled training data size is small. Our SRL system leverages unlabeled data under the language modeling paradigm. We demonstrate that the incorporation of a self pre-trained bidirectional language model (S-PrLM) into a SRL system can help in SRL performance improvement by learning composition functions from the unlabeled data. Previous researches have concluded that syntax information is very useful for high-performing SRL systems, so we incorporate syntax information by employing an unsupervised approach to leverage dependency path information to connect argument candidates in vector space, which helps in distinguishing arguments with similar contexts but different syntactic functions. The basic idea is to connect predicate ( w p ) with argument candidate ( w a ) with the dependency path ( r ) between them in the embedding space. Experiments on the CoNLL-2008 and CoNLL-2009 datasets confirm that our full SRL model outperforms previous best models in terms of F 1 score.

Non-Stationary Precipitation Frequency Estimates for Resilient Infrastructure Design in a Changing Climate: A Case Study in Sydney

The intensity–duration–frequency (IDF) curve is a commonly utilized tool for estimating extreme rainfall events that are used for many purposes including flood analysis. Extreme rainfall events are expected to become more intense under the changing climate, and there is a need to account for non-stationarity IDF curves to mitigate an underestimation of the risks associated with extreme rainfall events. Sydney, Australia, has recently started experiencing flooding under climate change and more intense rainfall events. This study evaluated the impact of climate change on altering the precipitation frequency estimates (PFs) used in generating IDF curves at Sydney Airport. Seven general circulation models (GCMs) were obtained, and the best models in terms of providing the extreme series were selected. The ensemble of the best models was used for comparing the projected 24 h PFs in 2031–2060 with historical values provided by Australian Rainfall and Runoff (ARR). The historical PFs consistently underestimate the projected 24 h PFs for all return periods. The projected 24 h 100 yr rainfall events are increased by 9% to 41% for the least and worst-case scenario compared to ARR historical PFs. These findings highlight the need for incorporating the impact of climate change on PFs and IDF curves in Sydney toward building a more prepared and resilient community. The findings of this study can also aid other communities in adapting the same framework for developing more robust and adaptive approaches to reducing extreme rainfall events’ repercussions under changing climates.

Batik Images Retrieval Using Pre-trained model and K-Nearest Neighbor

Batik is an Indonesian cultural heritage that should be preserved. Over time, many batik motifs have sprung up, which can lead to mutual claims between craftsmen. Therefore, it is necessary to create a system to measure the similarity of a batik motif. This research is focused on making Content-Based Image Retrieval (CBIR) on batik images. The dataset used in this research is big data Batik images. The authors used transfer learning on several pre-trained models and used Convolutional Neural Network (CNN) Autoencoder from previous studies to extract features on all images in the database. The extracted features calculate the Euclidean distance between the query and all images in the database to retrieve images. The image closest to the query will be retrieved according to the number of r, namely 3, 5, 10, or 15. Before the image is retrieved, the retrieval system is used to re-ranked with K-Nearest Neighbor (KNN), which classifies the retrieved image. The results of this study prove that MobileNetV2 + KNN is the best model in terms of Image Retrieval Batik, followed by InceptionV3 and VGG19 as the second and third ranks. Moreover, CNN Autoencoder from previous research and InceptionResNetV2 are ranked fourth and fifth. In this study, it was also found that the use of KNN re-ranking can increase the precision value by 0.00272. For further research, deploying these models, especially for MobileNetV2 is an approach for seeing a major impact on batik craftsmanship for decreasing batik motif plagiarism.

심전도 기반의 심장 질환 분류를 위한 신호 특징 추출 시스템 개발

In this paper, we propose a signal-based feature detection system for the early diagnosis of heart disease. The purpose of this study is to develop a compact healthcare system that extracts features from ECG signals. Therefore, the performance of the Tompkins algorithm, which is widely known as a signal-based feature detection algorithm, and the deep learning model of prior research are compared. In addition, by verifying the performance of the deep learning model by applying the Haar wavelet transform to the preprocessing, we determine the optimal feature detection model applicable to the healthcare system in terms of speed and accuracy. All algorithms and models were developed using Matlab on Window, and performance was compared on Jetson nano and Raspberry pi embedded boards. As a result, the best model in terms of speed and accuracy is the Haar wavelet bidirectional long short-term memory model, which quickly performs classification prediction with almost 97% accuracy. The advantage of this study is that it quickly performs classification prediction with high accuracy compared to existing healthcare systems that use algorithms that cause signal distortion. Therefore, cardiovascular diseases can be predicted and monitored based on the feature regions detected in this study. The data used are the MIT-BIH Arrhythmia Database and the Normal Sinus Rhythm Database. If additional patient data are established and verified in the future, it will be possible to use them not only in real life but also in clinical settings.

A Rank Graduation Accuracy measure to mitigate Artificial Intelligence risks

AbstractA key point to assess the applications of machine learning models in Artificial Intelligence (AI) is the evaluation of their predictive accuracy. This because the “automatic” choice of an action crucially depends on the made prediction. While the best model in terms of fit to the observed data can be chosen using a “universal” - and therefore automatable - criterion, based on the models’ likelihood, such as AIC and BIC, this is not the case for the best model in terms of predictive accuracy. To fill the gap, we propose a Rank Graduation Accuracy (RGA) measure which evaluates the concordance between the ranks of the predicted values and the ranks of the actual values of a series of observations to be predicted. We apply the RGA to a use-case that concerns the measurement of the financial risks that arise from crypto assets. The RGA appears as a “universal” alternative predictive model selection criterion that, differently from standard measures, such as the Root Mean Squared Error, is robust to the presence of outlying observations.

Relational Temporal Graph Reasoning for Dual-task Dialogue Language Understanding.

Dual-task dialog language understanding aims to tackle two correlative dialog language understanding tasks simultaneously via leveraging their inherent correlations. In this paper, we put forward a new framework, whose core is relational temporal graph reasoning. We propose a speaker-aware temporal graph (SATG) and a dual-task relational temporal graph (DRTG) to facilitate relational temporal modeling in dialog understanding and dual-task reasoning. Besides, different from previous works that only achieve implicit semantics-level interactions, we propose to model the explicit dependencies via integrating prediction-level interactions. To implement our framework, we first propose a novel model Dual-tAsk temporal Relational rEcurrent Reasoning network (DARER), which first generates the context-, speaker- and temporal-sensitive utterance representations through relational temporal modeling of SATG, then conducts recurrent dual-task relational temporal graph reasoning on DRTG, in which process the estimated label distributions act as key clues in prediction-level interactions. And the relational temporal modeling in DARER is achieved by relational graph convolutional networks (RGCNs). Then we further propose Relational Temporal Transformer (ReTeFormer), which achieves fine-grained relational temporal modeling via Relation- and Structure-aware Disentangled Multi-head Attention. Accordingly, we propose DARER with ReTeFormer (DARER2), which adopts two variants of ReTeFormer to achieve the relational temporal modeling of SATG and DTRG, respectively. The extensive experiments on different scenarios verify that our models outperform state-of-the-art models by a large margin. Remarkably, on the dialog sentiment classification task in the Mastodon dataset, DARER and DARER2 gain relative improvements of about 28% and 34% over the previous best model in terms of F1.

Path Loss Models for Cellular Mobile Networks Using Artificial Intelligence Technologies in Different Environments

One of the most critical problems in a communication system is losing information between the transmitter and the receiver. WiMAX (Worldwide Interoperability of Microwave Access) technology is gaining popularity and recognition as a Broadband Wireless Access (BWA) solution. At frequencies below 11 GHz, WiMAX can operate in line-of-sight (LOS) and non-line-of-sight (NLOS) scenarios. The implementation of WiMAX networks are rushed worldwide. Estimating path loss is crucial in the early stages of wireless network deployment and cell design. To anticipate propagation loss, several path loss models are available (e.g., Okumura Model Hata Model), but they are all bound by particular parameters. In this paper, we propose an MLP neural network-based path loss model with a well-structured implementation network design and grid search-based hyperparameter tuning. The proposed model optimally approximates mobile and base station path losses. Therefore, neurons number, learning rate, and hidden layers number are investigated to obtain the best model in terms of prediction accuracy. Path loss data is collected based on 14 networks in different microcellular settings. Simulations under Matlab environment showed that prediction errors were lower than standard log-distance-based path loss models.

Comparing Competing Systems: An Extension of the Information Systems Continuance Model

Although individual adoption and use of a single system has been examined extensively, little is known about how people evaluate and compare competing systems. In this paper, we discuss and test three alternative models underlying user comparison of competing systems: separate, crossover effect, and relative comparison processes. The separate comparison process proposes that users develop separate cognitive, affective, and conative evaluations toward each system, and the between-system comparison only occurs at the point of choosing a preferred system. The crossover effect comparison process posits that users not only perform separate evaluations for each system, but also consider the competitive effects when proceeding across cognitive, affective, and conative evaluation stages. In contrast, the relative comparison process postulates that users directly compare competing systems within each of the cognitive, affective, and conative evaluation stages. Based on the IS continuance model, we tested each of these three models using data collected from users of two competing instant messaging systems. Our results showed that the relative comparison process is the most parsimonious and the best model in terms of explaining the mechanisms underlying the comparison of system use by individuals. Theoretical and practical implications are discussed.

Impact of built environment on residential online car-hailing trips: Based on MGWR model.

With the development of smart mobile devices and global positioning technology, people's daily travel has become increasingly dependent on online car-hailing. Meanwhile, it has also become possible to use multi-source data to explore the factors influencing urban residents' car-hailing trips. Using online data on car-hailing trajectories, points of interest (POIs) data and other auxiliary data, the paper explores how the built environment impacts online car-hailing passengers. Within a 200 x 200m research grid, the unique spatiotemporal patterns of weekday car-hailing trips during a one-week period are analyzed, using statistics on pick-ups and drop-offs at different time of the day. By combining these data with built environment variables and various economic and traffic indicators, a multi-scale geographically weighted regression (MGWR) model is developed for different time scales. The MGWR model outperforms the classical geographically weighted regression (GWR) model and the ordinary least squares (OLS) regression model in terms of goodness of fit and all other aspects. More importantly, this study finds a high degree of temporal and spatial heterogeneity in the impact of built environment factors on local car-hailing trips across different regions, and the paper analyzes the business residence coefficient in detail. The study provides valuable insights to help improve the level of urban transportation services, as well as urban transportation planning and construction.

Abstract 14629: Predictive Model Testing for Life-Sustaining Treatment Decisions at the End-of-Life: Applying Machine Learning Methods

Introduction: Predictive models of factors affecting decisions about life-sustaining treatments (LSTs) have been frequently examined, but the accuracy and precision of those models remain elusive. Machine learning algorithms (MLAs) may help identify the best model in terms of its performance measures. Objectives: To determine the best MLA to predict LST preferences (cardiopulmonary resuscitation [CPR], ventilation support, and hemodialysis, and hospice care). Methods: In this secondary analysis, three MLA-based prediction models of LST preferences, including seven known predictors, were developed based on random forest (RF), logistics regression (LR), and linear support vector machine (SVM), and were evaluated by a five-fold validation set comprising accuracy, precision, area under the receiver operating characteristic curve, recall, and F1 score. Results: Among 375 adults with malignancy or non-malignancy diseases or healthy adults, 23.2% preferred CPR, 17.9% ventilation support, 19.2% hemodialysis, and 60.8% hospice care. Compared to RF and LR, SVM was the most accurate and precise model to predict each preference of CPR (0.78 vs. 0.85; 0.15-0.44 vs. 0.75, respectively), ventilation support (0.84 vs. 0.88; 0.00-0.10 vs. 0.89, respectively), hemodialysis (0.82 vs. 0.84; 0.33-0.54 vs. 0.69, respectively), and hospice care (0.64-0.68 vs. 0.85; 0.65-0.70 vs. 0.84, respectively). In the models, age, attitudes toward advance directives, and susceptibility to end-of-life care were reported as common predictors with relatively high importance on LST preferences (Table 1). Conclusions: SVM was the best model in identifying predictors of LST preferences in people with or without chronic diseases. In addition, the findings suggest important targets of interventions, including attitudes toward advance directives, and susceptibility to end-of-life care, to assist people to make the best decisions for their future medical care.

Machine Learning-Based Live Weight Estimation for Hanwoo Cow

Live weight monitoring is an important step in Hanwoo (Korean cow) livestock farming. Direct and indirect methods are two available approaches for measuring live weight of cows in husbandry. Recently, thanks to the advances of sensor technology, data processing, and Machine Learning algorithms, the indirect weight measurement has been become more popular. This study was conducted to explore and evaluate the feasibility of machine learning algorithms in estimating the body live weight of Hanwoo cow using ten body measurements as input features. Various supervised Machine Learning algorithms, including Multilayer Perceptron, k-Nearest Neighbor, Light Gradient Boosting Machine, TabNet, and FT-Transformer, are employed to develop the models that estimate the body live weight using body measurement data. Data analysis is exploited to explore the correlation between the body size measurements (the features) and the weights (target values that need to be estimated) of cows. Data analysis results show that ten body measurements have a high correlation with the body live weight. High performance of all applied Machine Learning models was obtained. It can be concluded that estimating the body live weight of Hanwoo cow is feasible by utilizing Machine Learning algorithms. Among all of the tested algorithms, LightGBM regression demonstrates not only the best model in terms of performance, model complexity and development time.

Research on adaptive combined wind speed prediction for each season based on improved gray relational analysis.

The stability of the power grid and the operational security of the power system depend on the precise prediction of wind speed. In consideration of the nonlinear and non-stationary characteristics of wind speed in different seasons, this paper employs the weight of wind resource index calculated by triangular fuzzy analytic hierarchy process (TF-AHP), criteria importance through inter-criteria correlation (CRITIC), and entropy weight method (EWM) to improve gray correlation analysis (GRA) and obtain the gray correlation degree of each season. In addition, a wind speed prediction model is provided that includes single-layer and two-layer weighting and is based on both deep and shallow machine learning models. At first, we establish each quarter's wind resource characteristics at typical monthly intervals of 10min, 30min, 60min, and 120min. The GRA's TF-AHP-CRITIC-EWM, enhanced with subjective and objective weights, is used to assess the available wind resources in each season and to compute the forecasted combination of wind speed for each season. As the final prediction results, the prediction values of each layer model are evaluated independently. For the intervals with considerable errors, we apply wavelet denoising and replacement combination. The simulation findings show that the proposed combined model surpasses earlier benchmark models in terms of goodness of fit, prediction accuracy, and generalizability.

The Prediction of Survival after Surgical Management of Bone Metastases of the Extremities-A Comparison of Prognostic Models.

Individualized survival prognostic models for symptomatic patients with appendicular metastatic bone disease are key to guiding clinical decision-making for the orthopedic surgeon. Several prognostic models have been developed in recent years; however, most orthopedic surgeons have not incorporated these models into routine practice. This is possibly due to uncertainty concerning their accuracy and the lack of comparison publications and recommendations. Our aim was to conduct a review and quality assessment of these models. A computerized literature search in MEDLINE, EMBASE and PubMed up to February 2022 was done, using keywords: “Bone metastasis”, “survival”, “extremity” and “prognosis”. We evaluated each model’s performance, assessing the estimated discriminative power and calibration accuracy for the analyzed patients. We included 11 studies out of the 1779 citations initially retrieved. The 11 studies included seven different models for estimating survival. Among externally validated survival prediction scores, PATHFx 3.0, 2013-SPRING and potentially Optimodel were found to be the best models in terms of performance. Currently, it is still a challenge to recommend any of the models as the standard for predicting survival for these patients. However, some models show better performance status and other quality characteristics. We recommend future, large, multicenter, prospective studies to compare between PATHfx 3.0, SPRING 2013 and OptiModel using the same external validation dataset.

Automatic Classification of Software Requirements in Vietnamese based on Machine Learning Techniques

Requirements engineering is often the first stage in the software process to understand the problem statement. Finding mistakes earlier in requirements helps reduce the development cost. One activity contributing to defining clear, complete and precise requirements is classifying requirement items in the specification. This paper presents a classification approach of functional and non-functional requirementsin Vietnamese using different supervised machine learning techniques. Five supervised machine learning algorithms, including Na¨ıve Bayes (NB), Support Vector Machine (SVM), Logistics Regression (LR), Multi-layer Perceptron Neural Network (MLP), and FastText, are implemented, trained, tested and compared using a dataset. The experimental results show that NB is the best model in terms of accuracy.

Comparison predicted using nonlinear model and in vivo studies growth production of broiler after fed different levels of fermented tamarind flour as replacing soya bean meal

This study compared two nonlinear models for growth performance of broiler after fed different levels of fermented tamarind flour as replacing soya bean meal. Two experiments were designed for this purpose. In first experiment, 20 chickens (Lohmann MB202) that had been fed on the different level of tamarind flour and weighed to determine best model suitable. In second experiment the treatment of the study was basal feed (control): basal feed substitute soy bean meal with tamarind flour 25%: basal feed substitute soy bean meal with tamarind flour 50%: basal feed substitute soy bean meal with tamarind flour 100%. The results showed that the used of tamarind flour did not give a significant effect (p > 0.05) on the production performance. It concluded the comparison of growth models in terms of goodness revealed that gompertz is the most appropriate model for describing between age and body weight in this experiment.

Model stacking to improve prediction and variable importance robustness for soft sensor development

This paper presents an interpretable ensemble modelling method, in which the predictions of several individual base learners are combined together through Stacked generalisation, which makes use of a secondary layer model, or so called meta-learner, that is trained on the output cross-validation predictions of each base learner. To provide interpretability, the permutation variable importance (PVI) is computed on the ensemble, wherein variables are randomly shuffled and the reduction in predictive performance for the ensemble is calculated for each variable. This is a novel contribution, as no previous attempts have been made in the soft sensor literature to investigate the interpretability of ensemble models that use heterogeneous base learners. The Stacked ensemble model also avoids model selection, which is the process of choosing among many candidate models. Model selection is often based on cross-validation, which is not guaranteed to select the best model in terms of true generalisation performance on the test set. Instead, the proposed method combines multiple models instead of choosing a singular model, avoiding the need for model selection. The efficacy of the proposed methodology in terms of both variable importance and predictive performance is shown on a synthetic dataset, in which the variable importance is already known, and an industrial dataset of a refinery process provided by Dow. For the synthetic dataset, it is shown that the proposed method chooses the correct casual variables, whereas the in-built variable importance provided by the individual models, namely Partial least squares, Lasso, Random forests & XGBoost, can give increased importance to non-causal, randomly generated variables. For the industrial study, the combined ensemble is shown to outperform all individual base models in terms of predictive performance, whilst also providing a new perspective in terms of variable importance compared to previous studies.