Search

Bloody Mahjong is a kind of mahjong. It is very popular in China in recent years. It not only has the characteristics of mahjong's conventional state space, huge hidden information, complicated rules, and large randomness of hand cards but also has special rules such as Change three, Hu must lack at least one suit, and Continue playing after Hu. These rules increase the difficulty of research. These special rules are used as the input of the deep learning DenseNet model. DenseNet is used to extract the Mahjong situation features. The learned features are used as the input of the classification algorithm XGBoost, and then the XGBoost algorithm is used to derive the card strategy. Experiments show that the fusion model of deep learning and XGBoost proposed in this paper has higher accuracy than the single model using only one of them in the case of high‐dimensional sparse features. In the case of fewer training rounds, accuracy of the model can still reach 83%. In the games against real people, it plays like human. [ABSTRACT FROM AUTHOR]

Background: Investigating emotion sequence patterns in the posts of discussion forums in massive open online courses (MOOCs) holds a vital role in shaping online interactions and impacting learning achievement. While the majority of research focuses on the relationship between emotions and interactions in MOOC forum discussions, research on identifying the crucial difference in emotion sequence patterns among different interaction groups remains in its infancy. Objectives: This research utilizes deep learning and sequential pattern mining to investigate whether there are differences in emotion sequence patterns across different groups of learners who exhibit various types of interactions in online discussion forums. Methods: Data from a comprehensive array of sources, including log files, discussion texts and scores from 498 learners in online discussion forums, were collected for this study. The agglomerative hierarchical algorithm is used to classify learners into groups with different levels of interactions. Additionally, we implement and evaluate multiple deep learning models for detecting different emotions from online discussions. Relevant emotion sequence patterns were identified using sequence pattern analysis and the identified emotion sequence patterns were compared across different groups with different levels of interactions. Results and Conclusions: Using an agglomerative hierarchical algorithm, we classified learners into three distinct groups characterized by different levels of interactions: high, average and low level. Leveraging the bi‐directional long short‐term memory model for emotion detection yielded the highest predictive performance, with an impressive F‐measure of 94.01%, a recall rate of 93.83% and an accuracy score of 95.01%. The results also revealed that learners in the low‐level interaction group experienced more emotion transition from boredom to frustration than the other two groups. Therefore, the aggregation of students into groups and the utilization of their MOOC log data offer educators the capability to provide adaptive emotional feedback, customize assessments and offer more personalized attention as needed. Lay Description: What is currently known about the subject matter: Emotions are dynamic over time when learners experience cognitive disequilibrium/equilibrium.Online interactions are critical components, which influence learners' emotional state, cognitive processes and learning achievement.It is not clear what are differences in emotion sequence patterns across the groups with different interaction types. What the paper adds: An agglomerative hierarchical algorithm was implemented to cluster learners into three groups by analysing behavioural data.We explore possibilities for automated classification of emotions using deep learning approaches.Learners in the low‐level interaction group experienced more emotion transition from boredom to frustration. Implications for practitioners: A considerable amount of effort should be expended to identify and respond to learners who experience boredom and frustration emotions.Designing interventions or scaffolding to facilitate learners' interaction and promote favourable emotions.Educators could provide more personalized support based on learners' online interaction cluster. [ABSTRACT FROM AUTHOR]

Background: Assignments that involve writing based on several texts are challenging to many learners. Formative feedback supporting learners in these tasks should be informed by the characteristics of evolving written product and by the characteristics of learning processes learners enacted while developing the product. However, formative feedback in writing tasks based on multiple texts has almost exclusively focused on essay product and rarely included SRL processes. Objectives: We explored the viability of using product and process features to develop machine learning classifiers that identify low‐ and high‐performing essays in a multi‐text writing task. Methods: We examined learning processes and essay submissions of 163 graduate students working on an authentic multi‐text writing assignment. We utilised learners' trace data to obtain process features and state‐of‐the‐art natural language processing methods to obtain product features for our classifiers. Results and Conclusions: Of four popular classifiers examined in this study, Random Forest achieved the best performance (accuracy = 0.80 and recall = 0.77). The analysis of important features identified in the Random Forest classification model revealed one product (coverage of reading topics) and three process (elaboration/organisation, re‐reading and planning) features as important predictors of writing quality. Major Takeaways: The classifier can be used as a part of a future automated writing evaluation system that will support at scale formative assessment in writing tasks based on multiple texts in different courses. Based on important predictors of essay performance, a guidance can be tailored to learners at the outset of a multi‐text writing task to help them do well in the task. Lay Description: What is already known about this topic?: Both product and process features should be used to inform formative feedback on writing.Providing product‐ and process‐oriented feedback to learners is challenging.Automatic writing evaluation systems have mainly relied upon product features.Automated analysis of learners' trace data and their essay drafts is a promising venue. What this paper adds?: An accurate machine learning classifier that identifies low‐ and high‐scoring essays.The classifier utilized both product and process features.We obtained process features from learners' trace data in digital learning environment.We computed product features using state‐of‐the‐art text analytical methods. Implications for practice and/or policy: The classifier can be used as a part of a future automated writing evaluation system.We revealed learning processes and essay characteristics that influence performance.Based on important predictors of performance, formative feedback can be given to learners. [ABSTRACT FROM AUTHOR]

While the use of experiments is important for developing students' scientific knowledge and skills, challenges may arise when teachers and students are conducting experiments in class, such as non‐reusable experimental resources, safety issues and difficulties simulating some specific effects. Augmented reality (AR) technology affords an alternative approach to conducting experiments by bringing students a virtual–real mixed learning environment. In this study, taking junior high school physics knowledge on the magnetic field as an example, we designed and developed an AR‐based mobile simulated experiment tool. This study investigates the effect of the AR‐based experiment on students' knowledge improvement and cognitive load compared with 3D and traditional experiments. A sample of 122 ninth‐grade students was randomly chosen and assigned to three groups (AR, 3D and Traditional). The results demonstrate that students in the AR group performed better than those in the 3D and Traditional groups in terms of their knowledge improvement. The AR group students also experienced the lowest cognitive load among the three groups. Moreover, students had positive perceptions about AR and 3D tools. The implications of this study are discussed. Lay Description: What is already known about this topic: Experiment is an important method for students' magnetic field learning.Traditional magnetic experiment consists of various procedures that may cause some inconvenience during learning.Augmented reality (AR) provides a virtual–real fusion environment to assist teaching and learning.While some studies indicate positive impact of AR on learner's cognitive load, some do not. What this paper adds: An AR‐based mobile simulated experiment tool was proposed based on the junior high school physics knowledge, the magnetic field.A Comparison was made among AR, 3D and traditional experiment for magnetic field learning.Higher knowledge improvement and lower cognitive load were observed in the AR group.3D and Traditional groups showed similar knowledge improvement and cognitive load. Implications for practice and/or policy: The AR design with hand‐held devices has unlimited scalability of educational applications.Serving the real object as marker and using the professional algorithm to simulate some specific phenomenon were recommended in future AR studies.The findings indicated that AR experimental tool had a statistically significant effect on students' knowledge improvement and cognitive load.The similar result between 3D and Traditional groups implies that not everything virtualized is good; we must consider the actual teaching and learning situations. [ABSTRACT FROM AUTHOR]

At present, China's Beidou Navigation Satellite System (BDS) has been successfully applied in all walks of life and entered a period of rapid development. The least squares algorithm is often used to solve problems in the application of BDS, while the quaternary nonlinear equation set is used for solving the BDS PVT positioning and timing issues. The solution of least squares is an iterative process, but if the convergence threshold is too small in the iterative process, it may lead to continuous iteration without convergence. Therefore, a least squares algorithm based on dynamic threshold was proposed in this paper. The proposed algorithm could continue iteration until convergence by increasing the system's dynamic threshold. At the same time, the proposed algorithm could also change the satellite combination in time, so as to conduct the positioning and calculation again over the threshold value that increases to a fixed value but fails to converge. Also, the BDS computation process was programmed and simulated. Simulation results show that even if the threshold is set too small, it converges well. During the operation of Beidou‐based geological disaster monitoring system in the Three Gorges Reservoir Region, the algorithm not only effectively prevents the program failure in Beidou satellite positioning caused by the convergence failure of algorithm iteration but also greatly improves the stability and reliability of the system program operation. [ABSTRACT FROM AUTHOR]

The projection of runoff in a large‐scale basin under climate change and human activities is crucial for the future management of water supplies. The impacts of climate change on runoff projections are associated with large uncertainties. In this study, using the optimization and selection of climate models, as well as an uncertainty analysis via the global Bayesian model averaging (BMA) approach, an ensemble projection framework was established to significantly improve the reliability of runoff projections in the future. The global BMA algorithm proposed in this study is an improvement on the BMA algorithm for large basins and is designed to reflect the differences among various models across the basin. In this algorithm, comprehensive BMA weights are obtained by considering the number (N) of stations in a basin. To verify the feasibility and improvement of this method, the runoff in 2050 and 2070 was projected with data from the Yellow River Basin, China. The runoff and its 90% confidence intervals at the six main stations in the Yellow River Basin were obtained. The increased evapotranspiration will exceed the increase in runoff generated by increase in precipitation in the future. The runoff in the upper and middle reaches of the Yellow River in 2050 and 2070 are projected to be 9% and 7% lower, respectively, than those in the reference period. The ensemble projection method proposed in this paper can be used as a widely applicable process in hydrometeorological ensemble projection and provides a basis for water resource management planning. Key Points: An ensemble projection framework was established to significantly improve the reliability of runoff in the futureThe global BMA algorithm is proposed to obtain comprehensive BMA weights by considering number of stations in a basinThe runoff of the Yellow River Basin in 2050 and 2070 will decrease by 9% and 7%, respectively [ABSTRACT FROM AUTHOR]

With increasing digital information availability, semantic web technologies have been employed to construct semantic digital libraries in order to ease information comprehension. The use of semantic web enables users to search or visualize resources in a semantic fashion. Semantic web generation is a key process in semantic digital library construction, which converts metadata of digital resources into semantic web data. Many text mining technologies, such as keyword extraction and clustering, have been proposed to generate semantic web data. However, one important type of metadata in publications, called affiliation, is hard to convert into semantic web data precisely because different authors, who have the same affiliation, often express the affiliation in different ways. To address this issue, this paper proposes a clustering method based on normalized compression distance for the purpose of affiliation disambiguation. The experimental results show that our method is able to identify different affiliations that denote the same institutes. The clustering results outperform the well-known k-means clustering method in terms of average precision, F-measure, entropy, and purity. [ABSTRACT FROM AUTHOR]

Accurate predictions of the air quality index (AQI) is critical for pollution control and air quality warning. However, this is challenging because of the nonlinearity of data and the uncertainty between data relationships. This paper proposes a combinatorial model based on an improved adaptive dynamic particle swarm optimization (ADPSO) algorithm to optimize a bidirectional gated recurrent unit (BiGRU) neural network to predict AQI time series and capture data dependence. The ADPSO method incorporates a dynamic spatial search strategy into the standard particle swarm optimization method, allowing the parameters to be dynamically adjusted to balance global and local search capabilities, thus improving the performance and effectiveness of this optimization process. Compared to the BiGRU model, the PSO‐BiGRU model, and the radial basis neural,the results of the improved algorithm reveal that the root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) of the ADPSO‐BiGRU predicted air pollution index are smaller than the errors of the other three models. The accuracy of the ADPSO‐BiGRU prediction model is higher than that of the other models, and it aids in the development of effective regional air quality management policies to reduce the negative impacts of pollution. [ABSTRACT FROM AUTHOR]

With the information overload of user-generated content in microblogging, users find it extremely challenging to browse and find valuable information in their first attempt. In this paper we propose a microblogging recommendation algorithm, TSI-MR (Topic-Level Social Influence-based Microblogging Recommendation), which can significantly improve users' microblogging experiences. The main innovation of this proposed algorithm is that we consider social influences and their indirect structural relationships, which are largely based on social status theory, from the topic level. The primary advantage of this approach is that it can build an accurate description of latent relationships between two users with weak connections, which can improve the performance of the model; furthermore, it can solve sparsity problems of training data to a certain extent. The realization of the model is mainly based on Factor Graph. We also applied a distributed strategy to further improve the efficiency of the model. Finally, we use data from Tencent Weibo, one of the most popular microblogging services in China, to evaluate our methods. The results show that incorporating social influence can improve microblogging performance considerably, and outperform the baseline methods. [ABSTRACT FROM AUTHOR]

This paper evaluated the MODerate resolution Imaging Spectroradiometer (MODIS) gross primary production (GPP) product (MOD17) by using estimated GPP from eddy-covariance flux measurements over an irrigated winter wheat and maize double-cropping field on the North China Plain in 2003–2004, and an alpine meadow on the Tibetan Plateau in 2002–2003. The mean annual GPP from MOD17 accounted for 1/2–2/3 of the surface estimated mean annual GPP for the alpine meadow, but only about 1/5–1/3 for the cropland. This underestimation was partly attributed to low estimates of leaf area index by a MODIS product (MOD15) because it is used to calculate absorbed photosynthetically active radiation in the MOD17 algorithm. The main reason is that the parameter maximum light use efficiency ( εmax) in the MOD17 algorithm was underestimated for the two biomes, especially for the cropland. Contrasted to the default, εmax was optimized using surface measurements. The optimized εmax for winter wheat, maize and meadow was 1.18, 1.81 and 0.73 g C/MJ, respectively. By using the surface measurements and optimized εmax , the MOD17 algorithm significantly improved the accuracy of GPP estimates. The optimum MOD17 algorithm explained about 82%, 68%, and 79% of GPP variance for winter wheat, maize, and meadow, respectively. These results suggest that it is necessary to adjust the MOD17 parameters for the estimation of cropland and meadow GPP, particularly over cropland. [ABSTRACT FROM AUTHOR]

Aim: To develop and validate models based on logistic regression and artificial intelligence for prognostic prediction of molar survival in periodontally affected patients. Materials and Methods: Clinical and radiographic data from four different centres across four continents (two in Europe, one in the United States, and one in China) including 515 patients and 3157 molars were collected and used to train and test different types of machine‐learning algorithms for their prognostic ability of molar loss over 10 years. The following models were trained: logistic regression, support vector machine, K‐nearest neighbours, decision tree, random forest, artificial neural network, gradient boosting, and naive Bayes. In addition, different models were aggregated by means of the ensembled stacking method. The primary outcome of the study was related to the prediction of overall molar loss (MLO) in patients after active periodontal treatment. Results: The general performance in the external validation settings (aggregating three cohorts) revealed that the ensembled model, which combined neural network and logistic regression, showed the best performance among the different models for the prediction of MLO with an area under the curve (AUC) = 0.726. The neural network model showed the best AUC of 0.724 for the prediction of periodontitis‐related molar loss. In addition, the ensembled model showed the best calibration performance. Conclusions: Through a multi‐centre collaboration, both prognostic models for the prediction of molar loss were developed and externally validated. The ensembled model showed the best performance in terms of both discrimination and validation, and it is made freely available to clinicians for widespread use in clinical practice. [ABSTRACT FROM AUTHOR]

Liver cancer has become the third cause that leads to the cancer death. For hepatocellular carcinoma (HCC), as the highly malignant type of liver cancer, its recurrence rate after operation is still very high because there is no reliable clinical data to provide better advice for patients after operation. To solve the challenging issue, in this work, we design a novel prediction model for recurrence of HCC using neighbor2vec based algorithm. It consists of three stages: (a) In the preparation stage, the Pearson correlation coefficient was used to explore the independent predictors of HCC recurrence, (b) due to the low correlation between individual dimension and prediction target, K‐nearest neighbors (KNN) were found as a K‐vectors list for each patient (neighbor2vec), (c) all vectors lists were applied as the input of machine learning methods such as logistic regression, KNN, decision tree, naive Bayes (NB), and deep neural network to establish the neighbor2vec based prediction model. From the experimental results on the real data from Shandong Provincial Hospital in China, the proposed neighbor2vec based prediction model outperforms all the other models. Especially, the NB model with neighbor2vec achieves up to 83.02, 82.86, 77.6%, in terms of accuracy, recall rates, and precision. This article is categorized under:Technologies > Data PreprocessingTechnologies > ClassificationTechnologies > Machine Learning [ABSTRACT FROM AUTHOR]

Secondary direct-current loss within a substation is a severe event. In China, each line protection device is generally powered by one set of secondary direct current (DC) power system within a substation of 110 kV voltage level. In this case, the line protection devices will be immediately taken out of service when the secondary DC is lost, and the lines protected by them will lose their corresponding protections. Once a fault occurs on the line, only the remote back-up protection can be relied on to isolate the fault, which expands the fault influence and increases its clearance time. To handle the above problem, a novel wide-area protection algorithm based on compensation voltage moduli comparison is proposed, which is independent of multi-terminal data synchronism. Then, a comprehensive protection strategy composed of the proposed protection algorithm and the original local distances protection is proposed. The results from a power systems computer aided design/electromagnetic transients including DC (PSCAD/EMTDC)-based simulation verify that the proposed protection algorithm has higher sensitivity. The comprehensive protection strategy could cope with most line faults effectively, even if only partial data are available, the selectivity and speed of which are better than that of distance protections. [ABSTRACT FROM AUTHOR]

While evapotranspiration (ET) is normally measured as one hydrologic component, evaporation (E), and transpiration (T) result from different physical‐biological processes. Using a two‐source model, a trapezoid framework has been widely applied in recent years. The key to applying the trapezoid framework model is the determination of the dry/wet boundaries of the land surface temperature‐fractional vegetation coverage trapezoid (LST‐fc). Although algorithms have been developed to characterize the two boundaries, there remains a significant uncertainty near the wet boundary which scatters in a discrete and uneven manner. It is therefore difficult to precisely locate the wet boundary. To address this problem, a Wet Boundary Algorithm (WBA) was developed in this study with the algorithm applied in the region of Huang‐Huai‐Hai plain of China, using the Pixel Component Arranging and Comparing Algorithm (PCACA) to retrieve ET from MODerate‐resolution Imaging Spectroradiometer (MODIS) Data. The eddy covariance (EC) measurements from Yucheng station was used to verify the modified model where the root mean square error (RMSE) of 17.8 W/m2, Bias of −7.2 W/m2 for latent heat flux (LE) simulation in 28 cloudless test days. The ratio of transpiration to evapotranspiration (T/ET) varied between 0.48 and 0.81 over the Huang‐Huai‐Hai plain. The spatial and temporal distribution of ET revealed that agriculture practices have a significant influence on the hydrological cycle, where crop growth promotes the magnitude of ET. Likewise, harvesting activities significantly reduce ET. The proposed WBA algorithm significantly reduces the uncertainty of the trapezoid ET model caused by wet edge positioning. The analysis of the impact of agricultural activities on ET provide a better understanding how human activities change the hydrological cycle at regional scales. [ABSTRACT FROM AUTHOR]

Contingency screening is vital for maintaining the transmission system reliability. Bilevel programming is a powerful tool for the accurate identification of critical events. To accelerate the contingency screening with bilevel programming, this study proposes a root-event-based dynamic space pruning algorithm. A root event is an event in which all the failed components have contributions to the event impact. Considering the fact that numerous N − k events describe the same critical scenario, the proposed space pruning principles partition the N − k event space into the subspace of root events and the complement subspace. To accelerate the screening of subspace of root events, bilevel programming is constructed to identify new critical events having high risk and causing more severe load shedding than the root event. Besides, relaxed bilevel programming is proposed to accelerate the screening of the complement subspace. The relaxed scheme replaces the constraints of power flow by constraints of transmission capacity, and enables the identification of most of the critical events in the complement subspace with a low computational cost. Case studies on the IEEE RTS and a provincial grid in China indicate that the proposed algorithm significantly accelerates the screening process without missing critical events. [ABSTRACT FROM AUTHOR]

Abstract: Background and purpose: Over the last 500 years, the most earthquakes with 10,000 or more fatalities and the most overall fatalities have occurred in China. Physical therapists must develop a better understanding of the patient and injury types that they are likely to treat post‐earthquake disasters. This systematic review of Chinese and Western literature identified the primary patients treated by physical therapists post‐earthquake disasters for injuries that negatively impacted physical function, activity, and participation. Methods: Comparisons were made between reports of earthquakes in China and reports from the rest of the world combined. Results: Sixty‐seven studies of 71,986 patients (51.8% male) at 40.6 ± 15 years of age were included. Studies were mostly prospective (n = 48, 71.6%). Reports of earthquakes in China represented more recently occurring disasters (p = .003) and more prospective research designs (p = .003). Reports from China also had a higher median fracture number (p = .004). Studies from China used manual muscle testing (p = .02), visual analogue pain scales (p = .008), Barthel index or modified Barthel index (p < .0001), and joint motion assessment (p = .007) with greater frequencies. Discussion: Physical therapists from China are more likely to treat patients with a fracture; however, physical therapists from both regions are likely to treat patients with general injuries representing poly‐trauma to multiple body regions, traumatic brain‐closed head injuries, spinal cord injuries, peripheral nerve injuries, and soft tissue injuries. Implications on physiotherapy practice: These data can help improve earthquake disaster planning, infrastructure development, and resource needs assessment effectiveness. More prospective research study designs and more recent earthquake disasters in China are likely associated with greater explicit use of valid and reliable outcome measurements such as joint motion assessment, manual muscle testing, visual analogue pain scale, and the Barthel index or modified Barthel index. [ABSTRACT FROM AUTHOR]

In recent years, millions of households have shifted from traditional TVs to smart TVs for viewing online videos on TV screens. In this article, we perform extensive analyses on a large‐scale online video viewing log on smart TVs. Because time influences almost every aspect of our lives, our aim is to understand temporal patterns of the online video viewing behavior of smart TV viewers at the crowd level. First, we measure the amount of time per hour spent in watching online videos on smart TV by each household on each day. By applying clustering techniques, we identify eight daily patterns whose peak hours occur in different segments of the day. The differences among households can be characterized by three types of temporal habits. We also uncover five periodic weekly patterns. There seems to be a circadian rhythm at the crow level. Further analysis confirms that there exists a holiday effect in the online video viewing behavior on smart TVs. Finally, we investigate the popularity variations of different video categories over the day. The obtained insights shed light on how we can partition a day to improve the performance of time‐aware video recommendations for smart TV viewers. [ABSTRACT FROM AUTHOR]

Accurate and precise wound measurements are a critical component of wound detection and assessment. Digital cameras are convenient and objective tools that are being increasingly used worldwide to assist with wound measurements and assessments. However, heterogeneous wounds and poor lighting conditions continue to be obstacles to wound area recognition. This study, therefore, provides an improved automated type-based wound area assessment method that is robust to lighting conditions and can distinguish between different wound tissues based on wound colors. The results of both laboratory and clinical applications of the proposed method show excellent consistency of manual area measurements. This proposed technology is expected to provide wound care specialists with more clinical information about heterogeneous wounds, thereby enabling prospective cost savings for therapy and treatment. [ABSTRACT FROM AUTHOR]

Evapotranspiration (ET) is an important factor for understanding hydrological processes and climate dynamics. Remote sensing is considered as the most promising tool for estimation of ET over a large spatial scale. The purpose of this study is to analyse and evaluate the MOD16 algorithm, developed by Nishida et al. (2003a,b), in winter wheat fields by using MODIS land products, MOD11-1and surface temperature and MOD13-standard normalized difference vegetation index (NDVI), as well as observations at the Yucheng Experimental Station, China in 2002. The original MOD16 is a dual-source ET model in which the evaporation fraction (EF) for bare soil surface is estimated using the VI-Ts diagram method, and EF for vegetation is determined by a function of canopy resistance, aerodynamic resistance and air temperature. We analysed a radiation budget sub-model and found that the estimate for solar radiation is acceptable only on cloud-free days. We also found that, in the original MOD 16, seasonal variations of minimum canopy resistance and physiological temperatures are not considered, which results in overestimation of canopy resistance when leaf area index (LAI) is less than 2.5. Because the strong unstable thermal stratification over the dry bare soil surface is ignored in the original algorithm, wind speed is overestimated and aerodynamic resistance above the vegetation surface is underestimated. We then modified the original MOD16 algorithm. The result shows that both EF and ET for vegetation estimated with the modified algorithm are in consistent with both the observations of an eddy covariance system and the calculations using the Penman-Monteith method. [ABSTRACT FROM AUTHOR]

Examines the potential applications of canonical correlation analysis (CCA) in regional science by describing its algorithm in a spatial context. Ability of CCA to deal with two sets of multivariate variables simultaneously and to produce structured and spatial meanings; Mutually independent relationship between transport and development in China's Zhujiang Delta.