Search

Objective: Subcutaneous Immunotherapy (SCIT) is the long-lasting causal treatment of allergic rhinitis (AR). How to enhance the adherence of patients to maximize the benefit of allergen immunotherapy (AIT) plays a crucial role in the management of AIT. This study aims to leverage novel machine learning models to precisely predict the risk of non-adherence of AR patients and related local symptom scores in 3 years SCIT. Methods: The research develops and analyzes two models, sequential latentvariable model (SLVM) of Stochastic Latent Actor-Critic (SLAC) and Long Short-Term Memory (LSTM). SLVM is a probabilistic model that captures the dynamics of patient adherence, while LSTM is a type of recurrent neural network designed to handle time-series data by maintaining long-term dependencies. These models were evaluated based on scoring and adherence prediction capabilities. Results: Excluding the biased samples at the first time step, the predictive adherence accuracy of the SLAC models is from 60% to 72%, and for LSTM models, it is 66%-84%, varying according to the time steps. The range of Root Mean Square Error (RMSE) for SLAC models is between 0.93 and 2.22, while for LSTM models it is between 1.09 and 1.77. Notably, these RMSEs are significantly lower than the random prediction error of 4.55. Conclusion: We creatively apply sequential models in the long-term management of SCIT with promising accuracy in the prediction of SCIT nonadherence in AR patients. While LSTM outperforms SLAC in adherence prediction, SLAC excels in score prediction for patients undergoing SCIT for AR. The state-action-based SLAC adds flexibility, presenting a novel and effective approach for managing long-term AIT. [ABSTRACT FROM AUTHOR]

Marked temporal point process models (MTPPs) aim to model event sequences and event markers (associated features) in continuous time. These models have been applied to various application domains where capturing event dynamics in continuous time is beneficial, such as education systems, social networks, and recommender systems. However, current MTPPs suffer from two major limitations, i.e., inefficient representation of event dynamic's influence on marker distribution and losing fine-grained representation of historical marker distributions in the modeling. Motivated by these limitations, we propose a novel model called Marked Point Processes with Memory-Enhanced Neural Networks (MoMENt) that can capture the bidirectional interrelations between markers and event dynamics while providing fine-grained marker representations. Specifically, MoMENt is constructed of two concurrent networks: Recurrent Activity Updater (RAU) to capture model event dynamics and Memory-Enhanced Marker Updater (MEMU) to represent markers. Both RAU and MEMU components are designed to update each other at every step to model the bidirectional influence of markers and event dynamics. To obtain a fine-grained representation of maker distributions, MEMU is devised with external memories that model detailed marker-level features with latent component vectors. Our extensive experiments on six real-world user interaction datasets demonstrate that MoMENt can accurately represent users' activity dynamics, boosting time, type, and marker predictions, as well as recommendation performance up to 76.5%, 65.6%, 77.2%, and 57.7%, respectively, compared to baseline approaches. Furthermore, our case studies show the effectiveness of MoMENt in providing meaningful and fine-grained interpretations of user-system relations over time, e.g., how user choices influence their future preferences in the recommendation domain. [ABSTRACT FROM AUTHOR]

Emphasizing the significance of Model Predictive Control (MPC) in modern optimization of control systems, the proposed research distinctively highlights its predictive prowess through the application of current state variables and well-structured mathematical models. We introduced a Predictive Current Control (PCC) strategy applied to a Three-Phase Inverter-fed Induction Motor (IM), with a particular focus on the Sequential Model methodology. The Sequential Model MPC algorithm employed a cost functional approach, predicated on the square of the discrepancy between reference and stator-measured currents of the IM in d-q reference frame. This method, implemented and tested in both MATLAB/Simulink and Python environments, utilized a minimization principle to guide the switching states of the inverter, thereby ensuring the accuracy of voltage signals for the induction motor. The projected study further included a comparative analysis of the electromagnetic torque, load currents, rotor speed, and angle deviations derived from the Sequential Model with those obtained through the Ant Colony Optimization (ACO) and Nelder-Mead methods. The results distinctly illustrated the robust adaptability of the Sequential Model methodology, outperforming the ACO and Nelder-Mead techniques in certain aspects such as minimum current errors, better speed regulations, and rotor angle trajectories.

With the growth of the internet, the landscape of commercial transactions has undergone significant changes. The digital revolution has permeated various aspects of our daily lives, notably transforming the financial system through digital money transfers across global distances. Digital transactions, including online purchases, money transfers, banking, and investing, have become routine. This shift has provided many advantages and opened the door for illegal activities such as money laundering. This research aims to identify transaction patterns to help algorithms recognize and flag scammers. We propose a deep learning-based sequential model framework to optimize fraud and non-fraud transaction classification. Proposed approach incorporates methods of ensemble learning including Gradient Boost, Random Forest, Logistic Regression, and Voting Classifiers, along with hyperparameter tuning to prevent overfitting. To solve the dataset inconsistency common in credit card dataset, we used techniques like under-sampling and the Synthetic Minority Over-sampling Technique (SMOTE) with specific machine learning (ML) algorithms. Our methodology involves data pre-processing, feature engineering, model selection, and evaluation. We leveraged the computational power of Google Colab for efficient model training and testing. Temporal features were extracted to capture transaction patterns over time, such as frequency and timing, along with features based on transaction amounts, merchant types, and geographic locations to detect subtle anomalies indicative of fraud. These features were input into a supervised learning algorithm, resulting in a model significantly outperforming baseline methods. The model was tested using a sizable dataset of anonymized transactions made with credit cards, and its effectiveness was measured using metrics such as accuracy, precision, recall, and F1-score. While it cannot entirely prevent fraudulent transactions, it effectively manages false positives through careful review, achieving an accuracy of 99.59%, surpassing other innovative models.

This paper presents the e-RULENet, which is a novel framework to train a data-driven model for remaining useful life estimation from long run-to-failure data with an end-to-end manner. In order to enable end-to-end learning, a change point from the healthy stage to the degradation stage is estimated for each instance from measurements. The change point estimation in the previous framework is computationally expensive since it utilizes entire measurements over the lifetime for each instance. The e-RULENet solves this issue with estimating the change points with segments sampled from run-to-failure data. It is evaluated on three datasets for RUL estimation including two real-world datasets, and on a dataset for tension estimation as another use case. It outperforms an approach without end-to-end learning on all the three datasets. The results indicates that the e-RULENet works not only for cases with multiple health stages but also for a single degradation stage, which does not have any changes points.

Emphasizing the significance of Model Predictive Control (MPC) in modern optimization of control systems, the proposed research distinctively highlights its predictive prowess through the application of current state variables and well-structured mathematical models. We introduced a Predictive Current Control (PCC) strategy applied to a Three-Phase Inverter-fed Induction Motor (IM), with a particular focus on the Sequential Model methodology. The Sequential Model MPC algorithm employed a cost functional approach, predicated on the square of the discrepancy between reference and stator-measured currents of the IM in d-q reference frame. This method, implemented and tested in both MATLAB/Simulink and Python environments, utilized a minimization principle to guide the switching states of the inverter, thereby ensuring the accuracy of voltage signals for the induction motor. The projected study further included a comparative analysis of the electromagnetic torque, load currents, rotor speed, and angle deviations derived from the Sequential Model with those obtained through the Ant Colony Optimization (ACO) and Nelder-Mead methods. The results distinctly illustrated the robust adaptability of the Sequential Model methodology, outperforming the ACO and Nelder-Mead techniques in certain aspects such as minimum current errors, better speed regulations, and rotor angle trajectories. [ABSTRACT FROM AUTHOR]

Objective: The sequential model has been defined as an intensive, two‐stage approach that comprises administering two types of treatment consecutively to improve treatment outcomes in cases of non‐optimal or absence of treatment response. A psychiatric population that would potentially benefit from the application of the sequential model is the eating disorders (EDs) population. The current scoping review aimed to explore the emerging literature on the application of sequential treatments in EDs. Method: Using PRISMA and Population intervention comparison outcomes study guidelines, Pubmed and PsycINFO were systematically searched for studies which applied temporally sequential treatments in patients diagnosed with EDs from inception to April 2022 using a combination of keywords. Studies utilising combined or integrated approaches were excluded. Results: A total of 12 studies were selected and reviewed. Studies included Bulimia Nervosa, Binge Eating Disorder (BED), or mixed ED samples with a majority of female patients. No studies on AN samples were identified. The majority of studies contained a Cognitive‐Behavioural Therapy module of treatment, were conducted on BED patients, were in outpatient settings, and included a group format in one or more treatment conditions. Studies varied in number of comparison groups and study design. Secondary and sequentially applied treatment modules were consistent with treatment recommendations of clinical guidelines. Conclusions: The available data on sequential treatments in EDs is scarce and exhibits methodological limitations that should be addressed in future studies. Definition of sequential treatments in EDs should be further developed to guide robust clinical research and improve empirical support of sequential treatment for complex ED cases and for non‐optimal ED treatment response. Highlights: Sequential treatments for eating disorders (EDs) have been applied exclusively in Bulimia Nervosa (BN), Binge Eating Disorder (BED) or mixed ED samples that are mostly female.Studies are heterogeneous in treatment paradigms, number of treatments applied, and comparison conditions, however, they are mostly psychological, based on Cognitive‐Behavioural Therapy (CBT) or on behavioural interventions.To date, studies are limited and lack proper experimental design including appropriate control groups to determine the utility of sequential treatment versus single treatment. [ABSTRACT FROM AUTHOR]

The increased availability of healthcare data has made predictive modeling popular in a clinical setting. If an expected patient-specific outcome can be estimated prior to a medical intervention the healthcare costs can be reduced for both patient and provider. The nature of data used to train such predictive models is frequently longitudinal, as interventions with convalescence times or chronic conditions contain outcome measures at intermediate follow-up points. Here we outline a predictive modeling approach that takes advantage of the longitudinal structure of the data by sequentially predicting the outcomes at intermediate time points and including them as predictors in models for later time points. This is done for continuous and threshold-dichotomized outcomes. The proposed method improves predictive accuracy as it takes advantage of the correlation in follow-up measures to distribute the estimation of coefficient effects over several models, making it advantageous for smaller datasets. This formulation also allows for effective screening of first-order interaction effects. The improved performance is illustrated using a simulation study and an applied example of predicting outcomes following surgery. The proposed approach is shown to be consistent for prediction, effective in modeling interactions and robust to presence of noise variables. [ABSTRACT FROM AUTHOR]

Climate change and global warming drive many governments and scientists to investigate new renewable and green energy sources. Special attention is on solar panel technology, since solar energy is considered one of the primary renewable sources and solar panels can be installed in domestic neighborhoods. Photovoltaic (PV) power prediction is essential to match supply and demand and ensure grid stability. However, the PV system has assertive stochastic behavior, requiring advanced forecasting methods, such as machine learning and deep learning, to predict day-ahead PV power accurately. Machine learning models need a rich historical dataset that includes years of PV power outputs to capture hidden patterns between essential variables to predict day-ahead PV power production accurately. Therefore, this study presents a framework based on the transfer learning method to use reliable trained deep learning models of old PV plants in newly installed PV plants in the same neighborhoods. The numerical results show the effectiveness of transfer learning in day-ahead PV prediction in newly established PV plants where a sizable historical dataset of them is unavailable. Among all nine models presented in this study, the LSTM models have better performance in PV power prediction. The new LSTM model using the inadequate dataset has 0.55 mean square error (MSE) and 47.07% weighted mean absolute percentage error (wMAPE), while the transferred LSTM model improves prediction accuracy to 0.168 MSE and 32.04% wMAPE.

Learning management systems (LMSs) have been used massively due to the growing utilization of distance learning. This advancement has led to increased educational data that can be analyzed to improve the quality of the learning process. Learning analytics (LA) is one of the most important methods that can be used to analyze student performance. In this paper, we proposed an LA method based on deep learning, i.e., transformer encoder, to sequentially predict the student’s final performance based on log activities provided by an LMS. The objective is to predict at-risk students of failing so that they can be mitigated as soon as possible. The proposed model was evaluated on the Open University LA Dataset (OULAD) for daily or weekly prediction. The results show that the model could predict at the early stage with an accuracy of 83.17% on withdrawn versus pass-distinction classes. Meanwhile, for the other tasks, i.e., withdrawn-fail versus pass-distinction and fail versus pass-distinction tasks, the accuracy was at least 76% at the early stage. The proposed model was compared to the LSTM model. We found that the transformer encoder performed better than the LSTM, with the average difference values from 1% to 3% in terms of accuracy and from 3% to 7% in terms of F1-score for all tasks, based on the statistical testing. Furthermore, the ablation study using positional encoding, different feature aggregation methods, and weighted loss function for the imbalanced class problem was conducted. In OULAD, we found that model without positional encoding was better in all cases. Furthermore, the weekly feature aggregation and the use of a weighted loss function performed better in some cases.

By using computer-aided arrhythmia diagnosis tools, electrocardiogram (ECG) signal plays a vital role in lowering the fatality rate associated with cardiovascular diseases (CVDs) and providing information about the patient's cardiac health to the specialist. Current advancements in deep-learning-based multivariate time series data analysis, such as ECG data classification include LSTM, Bi-LSTM, CNN, with Bi-LSTM, and other sequential networks. However, these networks often struggle to accurately determine the long-range dependencies among data instances, which can result in problems such as vanishing or exploding gradients for longer data sequences. To address these shortcomings of sequential models, a hybrid arrhythmia classification system using recurrence along with a self-attention mechanism is developed. This system utilizes convolutional layers as a part of representation learning, designed to capture the salient features of raw ECG data. Then, the latent embedded layer is fed to a self-attention-assisted transformer encoder model. Because the ECG data are highly influenced by absolute order, position, and proximity of time steps due to interdependent relationships among immediate neighbors, a component of recurrence using Bi-LSTM is added to the encoder model to address this characteristic of the data. The model performance indices such as classification accuracy and F1-score were found to be 99.2%. This indicates that the combination of recurrence along with self-attention-assisted architecture produces improved classification of arrhythmia from raw ECG signal when compared with the state-of-the-art models. [ABSTRACT FROM AUTHOR]

Climate change and global warming drive many governments and scientists to investigate new renewable and green energy sources. Special attention is on solar panel technology, since solar energy is considered one of the primary renewable sources and solar panels can be installed in domestic neighborhoods. Photovoltaic (PV) power prediction is essential to match supply and demand and ensure grid stability. However, the PV system has assertive stochastic behavior, requiring advanced forecasting methods, such as machine learning and deep learning, to predict day-ahead PV power accurately. Machine learning models need a rich historical dataset that includes years of PV power outputs to capture hidden patterns between essential variables to predict day-ahead PV power production accurately. Therefore, this study presents a framework based on the transfer learning method to use reliable trained deep learning models of old PV plants in newly installed PV plants in the same neighborhoods. The numerical results show the effectiveness of transfer learning in day-ahead PV prediction in newly established PV plants where a sizable historical dataset of them is unavailable. Among all nine models presented in this study, the LSTM models have better performance in PV power prediction. The new LSTM model using the inadequate dataset has 0.55 mean square error (MSE) and 47.07% weighted mean absolute percentage error (wMAPE), while the transferred LSTM model improves prediction accuracy to 0.168 MSE and 32.04% wMAPE. [ABSTRACT FROM AUTHOR]

We consider coupled models of flow and energy with phase change in thawing soils in permafrost at the pore- and Darcy scales. We develop the underlying models from first principles and define robust and conservative discretization algorithms based on cell-centered finite volume discretization on Cartesian grids and implicit time stepping for individual models. To couple the models we consider sequential approaches as well as time-lagging of some of the nonlinear properties. We also derive practical new surrogate models to obtain Darcy scale data from the pore-scale simulations. Finally, we study the sensitivity of models to the data in realistic scenarios. In particular we find that for the dynamics at the time scale of days, the nonlinear flow parameters have most significant impact on pressure, while temperature is less sensitive to data. [ABSTRACT FROM AUTHOR]

BackgroundA total of 11 treatment sequences for advanced wild-type squamous non-small cell lung cancer are recommended by Chinese Society of Clinical Oncology Guidelines, consisting of seven first-line and three second-line treatments. Five of these treatments were newly approved in China between 2021 and 2022. We evaluated the effectiveness and cost-effectiveness of these strategies from the Chinese healthcare system perspective.MethodsNetwork meta-analysis with non-proportional hazards was used to calculate the relative efficacy between interventions. A sequential model was developed to estimate costs and quality-adjusted life years (QALY) for treatment sequences with first-line platinum- and paclitaxel-based chemotherapy (SC) with or without nedaplatin, tislelizumab, camrelizumab, sintilimab, sugemalimab or pembrolizumab, followed by second-line docetaxel, tislelizumab or nivolumab. SC and docetaxel were used as comparators for first-line and second-line treatments, respectively. QALY and incremental cost-effectiveness ratio (ICER) were used to evaluate effectiveness and cost-effectiveness, respectively. Cost-effective threshold was set as USD 19,091. Subgroup analysis was conducted to determine the best first-line and second-line therapy.ResultsPembrolizumab + SC, followed by docetaxel (PED) was the most effective treatment sequence. QALYs for patients received SC, nedaplatin + SC, tislelizumab + SC, sintilimab + SC, camrelizumab + SC, sugemalimab + SC, pembrolizumab + SC followed by docetaxel were 0.866, 0.906, 1.179, 1.266, 1.179, 1.266, 1.603, 1.721, 1.807; QALYs for SC, nedaplatin + SC followed by tislelizumab were 1.283, 1.301; QALYs for SC, nedaplatin + SC followed by nivolumab were 1.353, 1.389. Camrelizumab + SC, followed by docetaxel (CAD) was the most cost-effective. Compared to SC with or without nedaplatin, tislelizumab, or sintilimab followed by docetaxel, ICERs of CAD were USD 12,276, 13,210, 6,974, 9,421/QALY, respectively. Compared with nedaplatin or SC followed by tislelizumab, the ICERs of CAD were USD 4,183, 2,804/QALY; CAD was dominant compared with nedaplatin or SC followed by nivolumab; The ICER of sugemalimab + SC followed by docetaxel and PED were USD 522,023, 481,639/QALY compared with CAD. Pembrolizumab + SC and camrelizumab + SC were the most effective and cost-effective first-line options, respectively; tislelizumab was the most effective and cost-effective second-line therapy. Tislelizumab used in second-line was more effective than first-line, no significant differences between their cost-effectiveness. Sensitivity and scenario analysis confirmed robustness of the results.ConclusionsPED and CAD are the most effective and cost-effective treatment sequence, respectively; pembrolizumab + SC and camrelizumab + SC are the most effective and cost-effective first-line choice, respectively; tislelizumab is the most effective and cost-effective second-line choice.

With the pervasiveness of secured biometric authentication applications, the fingerprint-based identification system has fascinated much attention recently. However, the major detriment is their recognition sensors are vulnerable to presentation or spoofing attacks from fake fingerprint artifacts. To resolve these issues, a viable anti-deception countermeasure known as presentation attack detection (PAD) mechanism is developed. As handcrafted feature-based classification techniques exhibit encouraging results in computer vision, they are widely employed in fingerprint spoof detection. Notably, the single-feature-based techniques do not perform uniformly over different spoofing and sensing technologies. In this research work, we expound a new hybrid fingerprint presentation attack detection approach (HyFiPAD) that discriminates live and fake fingerprints using majority voting ensemble build on three local and adaptive textural image features. We propose a new descriptor (i.e., a variant of LBP) which is termed as Local Adaptive Binary Pattern (LABP). Thus, the notion of proposed LABP is used to extract more detailed micro-textural features from the fingerprint images. Our LABP features are combined with an existing Complete Local Binary Pattern (CLBP) descriptor to learn two respective SVM classifiers and additionally a sequential model is trained with the manually extracted Binary Statistical Image Features (BSIF). The experiments are performed on benchmark anti-spoofing datasets namely; LivDet 2009, LivDet 2011, LivDet 2013, and LivDet 2015, where an average classification error rate (ACER) of 4.11, 3.19, 2.88, and 2.97% is, respectively, achieved. The overall experimental analysis of the HyFiPAD demonstrates superiority against majority of the state-of-the-art methods. In addition, the proposed technique yields a promising performance on cross-database and cross-sensor liveness detection tests, claiming good generalization capability. [ABSTRACT FROM AUTHOR]

Growth trends in passenger transport demand and gross domestic product have so far been similar. The increase in mobility in one area is connected with the increase in GDP in the same area. This increase is representative of the economic and social development of the area. At the same time, the increase in mobility produces one of the most negative environmental impacts, mainly determined by the growth of mobility of private cars. International attention is given to the possibilities of increasing mobility and, therefore, social and economic development without increasing environmental impacts. One of the most promising fields is that of MaaS: Mobility as a Service. MaaS arises from the interaction of new user behavioral models (demand) and new decision-making models on services (supply). Advanced interaction arises from the potentialities allowed by emerging ICT technologies. There is a delay in the advancement of transport system models that consider the updating of utility and choice for the user by means of updated information. The paper introduces sustainability as defined by Agenda 2030 with respect to urban passenger transport, then examines the role of ICT in the development of MaaS formalizing a dynamic model of demand–supply interaction explicating ICT. Finally, the advanced Sustainable MaaS, defined SMaaS, is analyzed, evidencing the contribution to achieving the goals of Agenda 2030. [ABSTRACT FROM AUTHOR]

In cognitive diagnostic assessments with time limits, not-reached items (i.e., continuous nonresponses at the end of tests) frequently occur because examinees drop out of the test due to insufficient time. Oftentimes, the not-reached items are related to examinees' specific cognitive attributes or knowledge structures. Thus, the underlying missing data mechanism of not-reached items is non-ignorable. In this study, a missing data model for not-reached items in cognitive diagnosis assessments was proposed. A sequential model with linear restrictions on item parameters for missing indicators was adopted; meanwhile, the deterministic inputs, noisy "and" gate model was used to model the responses. The higher-order structure was used to capture the correlation between higher-order ability parameters and dropping-out propensity parameters. A Bayesian Markov chain Monte Carlo method was used to estimate the model parameters. The simulation results showed that the proposed model improved diagnostic feedback results and produced accurate item parameters when the missing data mechanism was non-ignorable. The applicability of our model was demonstrated using a dataset from the Program for International Student Assessment 2018 computer-based mathematics cognitive test. [ABSTRACT FROM AUTHOR]

In many real-world decision-making tasks, multi-agent need to learn collaboration in a high-dimensional complex action space, rather than just a single discrete action space. Recently, value decomposition learning methods such as QMIX have emerged as a promising approaches for collaborative multi-agent tasks. However, most of the value decomposition algorithms can only be used in discrete action space, which would limit their practicability. To address the limitation, we propose a novel algorithm called Multi-Agent Sequential Q-Networks (MASQN), which can be applied to the multi-agent domains with continuous, multidiscrete or hybrid action spaces. The proposed algorithm is based on the structure of centralized training with decentralized execution (CTDE). The decentralized actors ensure adaptability to different action spaces by utilizing action space discretization and sequential models, and the centralized critic utilizes the value decomposition architecture to guide effective updates of the policy parameters for each agent. We also give the convergence of joint policy from the perspective of policy iteration, by combining it with the CTDE structure and the constraint of the Individual Global Max (IGM) condition. Finally, we evaluate the MASQN algorithm on two benchmark environments: MAMuJoCo and Hybrid Predator–Prey. The empirical results show that MASQN out performs the state-of-the-art performance on three different action spaces. • This MARL algorithm can be directly used in continuous action spaces. • This MARL algorithm can be directly used in multi-discrete action spaces. • This MARL algorithm can be directly used in hybrid action spaces. • We analyzed the convergence of this algorithm in our paper. [ABSTRACT FROM AUTHOR]

Dengue fever is among the most dangerous infectious viral diseases transmitted through the bite of infected Aedes Aegypti mosquitoes. One way to decline the spread of dengue is by raising awareness to the community about mosquito habitats through continuous surveillance. The traditional surveillance techniques of Aedes Aegypti are difficult, time taking, and can lead to severe health risks. This paper presents a possible way of dengue vector surveillance through acoustic signals generated by wingbeat of Aedes Aegypti using the sequential model of convolutional neural network. Mel-frequency spectrum is given as an input feature to the sequential model that significantly improves classification performance up to 93% accuracy. The system generates notification through a specially designed mobile application to alert detected dengue vectors in the region. It is helpful in continuous monitoring of dengue vectors to take early precautionary measures for effective control and prevention.

In cognitive diagnostic assessments with time limits, not-reached items (i.e., continuous nonresponses at the end of tests) frequently occur because examinees drop out of the test due to insufficient time. Oftentimes, the not-reached items are related to examinees’ specific cognitive attributes or knowledge structures. Thus, the underlying missing data mechanism of not-reached items is non-ignorable. In this study, a missing data model for not-reached items in cognitive diagnosis assessments was proposed. A sequential model with linear restrictions on item parameters for missing indicators was adopted; meanwhile, the deterministic inputs, noisy “and” gate model was used to model the responses. The higher-order structure was used to capture the correlation between higher-order ability parameters and dropping-out propensity parameters. A Bayesian Markov chain Monte Carlo method was used to estimate the model parameters. The simulation results showed that the proposed model improved diagnostic feedback results and produced accurate item parameters when the missing data mechanism was non-ignorable. The applicability of our model was demonstrated using a dataset from the Program for International Student Assessment 2018 computer-based mathematics cognitive test.

By using computer-aided arrhythmia diagnosis tools, electrocardiogram (ECG) signal plays a vital role in lowering the fatality rate associated with cardiovascular diseases (CVDs) and providing information about the patient’s cardiac health to the specialist. Current advancements in deep-learning-based multivariate time series data analysis, such as ECG data classification include LSTM, Bi-LSTM, CNN, with Bi-LSTM, and other sequential networks. However, these networks often struggle to accurately determine the long-range dependencies among data instances, which can result in problems such as vanishing or exploding gradients for longer data sequences. To address these shortcomings of sequential models, a hybrid arrhythmia classification system using recurrence along with a self-attention mechanism is developed. This system utilizes convolutional layers as a part of representation learning, designed to capture the salient features of raw ECG data. Then, the latent embedded layer is fed to a self-attention-assisted transformer encoder model. Because the ECG data are highly influenced by absolute order, position, and proximity of time steps due to interdependent relationships among immediate neighbors, a component of recurrence using Bi-LSTM is added to the encoder model to address this characteristic of the data. The model performance indices such as classification accuracy and F1-score were found to be 99.2%. This indicates that the combination of recurrence along with self-attention-assisted architecture produces improved classification of arrhythmia from raw ECG signal when compared with the state-of-the-art models.

Background: Influenza vaccination efficacy is reduced after hematopoietic stem cell transplantation (HSCT) and patient factors determining vaccination outcomes are still poorly understood.Methods: We investigated the antibody response to seasonal influenza vaccination in 135 HSCT patients and 69 healthy volunteers (HVs) in a prospective observational multicenter cohort study. We identified patient factors associated with hemagglutination inhibition titers against A/California/2009/H1N1, A/Texas/2012/H3N2, and B/Massachusetts/2012 by multivariable regression on the observed titer levels and on seroconversion/seroprotection categories for comparison.Results: Both regression approaches yielded consistent results but regression on titers estimated associations with higher precision. HSCT patients required 2 vaccine doses to achieve average responses comparable to a single dose in HVs. Prevaccination titers were positively associated with time after transplantation, confirming that HSCT patients can elicit potent antibody responses. However, an unrelated donor, absolute lymphocyte counts below the normal range, and treatment with calcineurin inhibitors lowered the odds of responding.Conclusions: HSCT patients show a highly heterogeneous vaccine response but, overall, patients benefited from the booster shot and can acquire seroprotective antibodies over the years after transplantation. Several common patient factors lower the odds of responding, urging identification of additional preventive strategies in the poorly responding groups.Clinical Trials Registration: NCT03467074. [ABSTRACT FROM AUTHOR]

Ordinal models can be seen as being composed from simpler, in particular binary models. This view on ordinal models allows to derive a taxonomy of models that includes basic ordinal regression models, models with more complex parameterizations, the class of hierarchically structured models, and the more recently developed finite mixture models. The structured overview that is given covers existing models and shows how models can be extended to account for further effects of explanatory variables. Particular attention is given to the modeling of additional heterogeneity as, for example, dispersion effects. The modeling is embedded into the framework of response styles and the exact meaning of heterogeneity terms in ordinal models is investigated. It is shown that the meaning of terms is crucially determined by the type of model that is used. Moreover, it is demonstrated how models with a complex category‐specific effect structure can be simplified to obtain simpler models that fit sufficiently well. The fitting of models is illustrated by use of a real data set, and a short overview of existing software is given. This article is categorized under:Statistical Models > Fitting ModelsData: Types and Structure > Categorical DataStatistical Models > Generalized Linear Models [ABSTRACT FROM AUTHOR]

Objective: Subcutaneous Immunotherapy (SCIT) is the long-lasting causal treatment of allergic rhinitis (AR). How to enhance the adherence of patients to maximize the benefit of allergen immunotherapy (AIT) plays a crucial role in the management of AIT. This study aims to leverage novel machine learning models to precisely predict the risk of non-adherence of AR patients and related local symptom scores in 3 years SCIT., Methods: The research develops and analyzes two models, sequential latent-variable model (SLVM) of Stochastic Latent Actor-Critic (SLAC) and Long Short-Term Memory (LSTM). SLVM is a probabilistic model that captures the dynamics of patient adherence, while LSTM is a type of recurrent neural network designed to handle time-series data by maintaining long-term dependencies. These models were evaluated based on scoring and adherence prediction capabilities., Results: Excluding the biased samples at the first time step, the predictive adherence accuracy of the SLAC models is from 60% to 72%, and for LSTM models, it is 66%-84%, varying according to the time steps. The range of Root Mean Square Error (RMSE) for SLAC models is between 0.93 and 2.22, while for LSTM models it is between 1.09 and 1.77. Notably, these RMSEs are significantly lower than the random prediction error of 4.55., Conclusion: We creatively apply sequential models in the long-term management of SCIT with promising accuracy in the prediction of SCIT nonadherence in AR patients. While LSTM outperforms SLAC in adherence prediction, SLAC excels in score prediction for patients undergoing SCIT for AR. The state-action-based SLAC adds flexibility, presenting a novel and effective approach for managing long-term AIT., Competing Interests: Authors PV and NC were employed by company Volkswagen Group. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Li, Xiong, Fan, Wang, Yu, Si, van der Smagt, Tang and Chen.)

Collaborative Filtering (CF)-based recommendation methods suffer from (i) sparsity (have low user–item interactions) and (ii) cold start (an item cannot be recommended if no ratings exist). Systems using clustering and pattern mining (frequent and sequential) with similarity measures between clicks and purchases for next-item recommendation cannot perform well when the matrix is sparse, due to rapid increase in number of items. Additionally, they suffer from: (i) lack of personalization: patterns are not targeted for a specific customer and (ii) lack of semantics among recommended items: they can only recommend items that exist as a result of a matching rule generated from frequent sequential purchase pattern(s). To better understand users' preferences and to infer the inherent meaning of items, this paper proposes a method to explore semantic associations between items obtained by utilizing item (products') metadata such as title, description and brand based on their semantic context (co-purchased and co-reviewed products). The semantics of these interactions will be obtained through distributional hypothesis, which learns an item's representation by analyzing the context (neighborhood) in which it is used. The idea is that items co-occurring in a context are likely to be semantically similar to each other (e.g., items in a user purchase sequence). The semantics are then integrated into different phases of recommendation process such as (i) preprocessing, to learn associations between items, (ii) candidate generation, while mining sequential patterns and in collaborative filtering to select top-N neighbors and (iii) output (recommendation). Experiments performed on publically available E-commerce data set show that the proposed model performed well and reflected user preferences by recommending semantically similar and sequential products. [ABSTRACT FROM AUTHOR]

In group decision support systems, it is important on how to process and manage individual decision information. In the paper, a sequential model is proposed to manage individual judgements with additively reciprocal property over paired alternatives. The process of realizing additive complementary pairwise comparisons (ACPCs) is captured. A real-time feedback mechanism is constructed to address the irrational behavior of individuals. An optimization model is established and solved by using the particle swarm optimization (PSO) algorithm, such that the consistency of individual judgements can be improved fast yet effectively. For the aggregation of individual decision information in group decision making (GDM), the weighted averaging operator is used. It is found that when all individual judgements are acceptably additively consistent, the collective matrix is with acceptable additive consistency. Under the control of individual consistency degrees, the approach of reaching consensus in GDM is further proposed. By comparing with some existing models, the observations reveal that the sequential model of originating additive complementary pairwise comparisons possesses the ability to rationally manage individual decision information. [ABSTRACT FROM AUTHOR]

Development of Intelligent Conversational Agent using Artificial Intelligence and Machine Learning technique is an interesting problem in the field of natural language processing (NLP). Conversational agents, widely known as chatbots, are being used as virtual assistant in many different domains including business, healthcare and government organizations in attempt to improve the quality of service provided. This paper aims at using NLP towards the development of chatbot to assist the new research scholars. New research scholars often find it difficult where to start, how to start and have some common queries about fundamental concepts in research, research publications, data sources, funding agencies, indexing services etc. The paper addresses the need of virtual assistant to the new researcher scholars and describes the design of the chatbot model that would help them get the answers to the research queries at their fingertips. The database used in the study is generated by collecting data from web as well as through personal interactions with new research scholars. We have designed retrieval based text chatbot using Keras sequential model with Adagrad optimizer that is optimized for learning rate. The research also reveals the relationship between learning rate and accuracy for five optimizers. It was found that accuracy of Adam and RMSprop optimizers is inversely proportional to the learning rate whereas the accuracy of Adadelta optimizer is directly proportional to the learning rate. The model designed in this work is simple, computationally inexpensive and has succeeded to provide answers to user queries satisfactorily. The paper gives important directions in which research can further be extended. [ABSTRACT FROM AUTHOR]

Summary: Appropriate modelling of Likert‐type items should account for the scale level and the specific role of the neutral middle category, which is present in most Likert‐type items that are in common use. Powerful hierarchical models that account for both aspects are proposed. To avoid biased estimates, the models separate the neutral category when modelling the effects of explanatory variables on the outcome. The main model that is propagated uses binary response models as building blocks in a hierarchical way. It has the advantage that it can be easily extended to include response style effects and non‐linear smooth effects of explanatory variables. By simple transformation of the data, available software for binary response variables can be used to fit the model. The proposed hierarchical model can be used to investigate the effects of covariates on single Likert‐type items and also for the analysis of a combination of items. For both cases, estimation tools are provided. The usefulness of the approach is illustrated by applying the methodology to a large data set. [ABSTRACT FROM AUTHOR]

When authenticating illegal users, the existing account/password matching mechanism is only used to validate passwords. Therefore, it can not identify the web behavior of malicious users, such as identity fraud. In the case of fingerprint and face recognition mechanisms, which are prone to fraud, these mechanisms are still ineffective in preventing identity fraud. Recent years, behavioral analysis has been widely studied in view of the fraud-resistant characteristics of user behavior data and the personalized information contained therein. Among them, mobile terminal behavior is one of the main types of web user behavior. Aiming at these mobile terminal behaviors, a method of subspace dimension reduction is proposed to transform the multi-variable real-valued time series into univariable. It is realized by forming each sliding window of multivariate time series into a matrix. By this it generates a matrix sequence. Then we compare the angle difference between two adjacent matrices and transform the matrix sequence into a univariate time series, and then we propose the Gaussian-processes based authentication method to verify a user's web behavior. The whole process is called Subspace-based dimension Reduction Gaussian-process Authentication (SRGA) method. Finally, the effectiveness of proposed method is verified by a series of experiments. [ABSTRACT FROM AUTHOR]

Autonomous systems with nonlinear dynamics need to be extremely resilient to errors in sensors, actuators and on-board electronics for the purpose of overall vehicle safety. Prior research has focused on control-theoretic methods with significant computational burden with a focus on failures in actuation. In contrast, we propose the use of hierarchical machine learning driven state space checks that detect and diagnose errors in control program execution, sensors and actuators with high sensitivity and low latency. Each check produces a time-varying error signal that facilitates effect-cause diagnosis of the system, while allowing rapid parameter estimation from each check. Since the checks are over small subsets of system parameters, estimation is fast and accurate. The estimated parameters are then used to reconfigure the system controller parameters for rapid system recovery. We use a quadcopter system to demonstrate and validate our approach. Controller, sensor and actuator errors can be detected, diagnosed and compensated using a common checking platform with low computational overhead. The technique is validated on a quadcopter hardware test vehicle. [ABSTRACT FROM AUTHOR]

Poultry farms across the world house animals such as cows, sheep's, pigs and hen These farms are the places from where the market gets meat, eggs, wool and other animal products which are used in our daily lives. But sometimes the business of the farms is severely hit primarily due to poultry animals catching diseases or these animals succumbing to injuries caused due to fighting each other. The identification of such animals showing unusual characteristics or behavior is necessary. We through our paper use deep learning concepts to help the poultry owner to identify these unusual characteristics in a sheep. Some of the characteristics that our paper identifies are skinny, redness, non-aggression, aggression, beefy and foraging. Our paper makes use of a video, shot from a camera and the implementation of a sequential model as well as make use of the SSD algorithm to identify and characterize the sheep's depicted in the video. The sequential model is trained using the training dataset which contains static images of sheep's and the attribute/characteristics depicted by those sheep's that are stored in a CSV file. The testing dataset contains the images that are extracted from the input video as frames. The testing dataset is passed through the sequential model to get the characteristics/attributes depicted by the sheep in each frame and store those characteristics/attributes in a CSV file. The SSD algorithm is trained on identifying the various animals and not only does it display the name of the animal detected, it displays the confidence percentage of the animal as well, in our paper it is used for identifying sheep. The above stated algorithm also creates a border around the identified sheep such that it can be used for tracking purposes during the entirety of the video. The SSD algorithm also takes in the attributes depicted by these frames and displays them along the border which identifies the sheep. The accuracy is compared with YOLO algorithms and shows 3 to 6% improvement on prediction rate. [ABSTRACT FROM AUTHOR]

The normativity of workers' actions during producing has a great impact on the quality of the products and the safety of the operation process. Previous studies mainly focused on the normativity of each single producing action instead of considering the normativity of continuous producing actions, which is defined as producing action flow (PAF) in this paper, during operation process. For this issue, a normativity judging method based on two-LSTM fusion network (TFN) and normativity-aware attention network (NAN) is proposed. First, TFN is designed to detect and recognize the producing actions based on skeleton sequences of a worker during complete operation process, and PAF data in sequential form are obtained. Then, NAN is built to allocate different levels of attention to each producing action within the sequence of PAF, and by this means, an efficient normativity judging is conducted. The combustor surface cleaning (CSC) process of rocket engine is taken as the experimental case, and the CSC-Action2D dataset is established for evaluation. Experiment results show the high performance of TFN and NAN, demonstrating the effectiveness of the proposed method for PAF normativity judging. [ABSTRACT FROM AUTHOR]