Predicting the displacement of landslide is of utmost practical importance as the landslide can pose serious threats to both human life and property.However,traditional methods have the limitation of random selection ...Predicting the displacement of landslide is of utmost practical importance as the landslide can pose serious threats to both human life and property.However,traditional methods have the limitation of random selection in sliding window selection and seldom incorporate weather forecast data for displacement prediction,while a single structural model cannot handle input sequences of different lengths at the same time.In order to solve these limitations,in this study,a new approach is proposed that utilizes weather forecast data and incorporates the maximum information coefficient(MIC),long short-term memory network(LSTM),and attention mechanism to establish a teacher-student coupling model with parallel structure for short-term landslide displacement prediction.Through MIC,a suitable input sequence length is selected for the LSTM model.To investigate the influence of rainfall on landslides during different seasons,a parallel teacher-student coupling model is developed that is able to learn sequential information from various time series of different lengths.The teacher model learns sequence information from rainfall intensity time series while incorporating reliable short-term weather forecast data from platforms such as China Meteorological Administration(CMA)and Reliable Prognosis(https://rp5.ru)to improve the model’s expression capability,and the student model learns sequence information from other time series.An attention module is then designed to integrate different sequence information to derive a context vector,representing seasonal temporal attention mode.Finally,the predicted displacement is obtained through a linear layer.The proposed method demonstrates superior prediction accuracies,surpassing those of the support vector machine(SVM),LSTM,recurrent neural network(RNN),temporal convolutional network(TCN),and LSTM-Attention models.It achieves a mean absolute error(MAE)of 0.072 mm,root mean square error(RMSE)of 0.096 mm,and pearson correlation coefficients(PCCS)of 0.85.Additionally,it exhibits enhanced prediction stability and interpretability,rendering it an indispensable tool for landslide disaster prevention and mitigation.展开更多
This paper presents the use of a student model to improve the explanations provided by an intelligent tu- toring system,namely SimpleQuestl,in the domain of electronics.The method of overlay modelling is adopted to bu...This paper presents the use of a student model to improve the explanations provided by an intelligent tu- toring system,namely SimpleQuestl,in the domain of electronics.The method of overlay modelling is adopted to build the student model.The diagnosis is based on the comparison of the behavinurs of the student and the ex- pert.The student model is consulted by the “explainer” and “debugging” procedures in order to re-order the sequence of the explanation.展开更多
It focuses on that students must be developed the ability to solve the practical problem by building the mathematics models and the ability to combine the theory with the practice. It also states that students must be...It focuses on that students must be developed the ability to solve the practical problem by building the mathematics models and the ability to combine the theory with the practice. It also states that students must be improved the learning interests and practical experience.展开更多
In this study, the mathematical models of dynamics of student populations in the university departments are formulated. As a case study, we employ the data of registration section from Department of Mathematics, Facul...In this study, the mathematical models of dynamics of student populations in the university departments are formulated. As a case study, we employ the data of registration section from Department of Mathematics, Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok (KMUTNB), Thailand, from calendar year 2006 to 2010. Using regression analysis, descriptive model and explanatory model are derived. The descriptive model is linear with R2 = 0.8864. Using log-transformation, the explanatory model gives the nonlinear approximation with R2 = 0.8293. The model predicts that the number of students of Department of Mathematics, KMUTNB has a tendency to linearly increase with slope of 20 with 95% CI (6.8417, 33.1583). The application of the models in educational management is discussed.展开更多
针对基于知识蒸馏的工业图像异常检测中,预训练网络在迁移到工业图像领域时因领域差异和数据分布的不同而导致获取的特征存在偏差问题,提出了一种特征自适应师生模型。该模型利用特征自适应器来调整教师网络提取的预训练特征,实现跨域...针对基于知识蒸馏的工业图像异常检测中,预训练网络在迁移到工业图像领域时因领域差异和数据分布的不同而导致获取的特征存在偏差问题,提出了一种特征自适应师生模型。该模型利用特征自适应器来调整教师网络提取的预训练特征,实现跨域特征转换和减少领域偏差。为了避免特征自适应器过度调整预训练特征,导致与原始预训练特征差异过大而降低泛化性能,提出监督损失来约束调整后的特征。此外,进一步为了提高模型对异常特征的表征和判别能力,设计了一个新的学生网络和提出了一种对抗损失来拉远教师和学生网络中异常特征之间的差异,拉近两者正常特征之间的差异。在多个工业数据集MVTec AD、BTAD、VisA和MVTec 3D AD上验证了该模型的有效性。展开更多
知识追踪根据学生的历史学习轨迹来实现对其知识掌握水平的实时评估与跟踪,从而预测学生未来的学习表现,是在线学习系统实现个性化学习体验的核心技术之一.与传统方法相比,现有的基于深度神经网络的知识追踪模型取得了显著优势,但其通...知识追踪根据学生的历史学习轨迹来实现对其知识掌握水平的实时评估与跟踪,从而预测学生未来的学习表现,是在线学习系统实现个性化学习体验的核心技术之一.与传统方法相比,现有的基于深度神经网络的知识追踪模型取得了显著优势,但其通常依赖大量训练数据.在学生答题早期,交互数据极度稀缺,所以训练一个复杂、有效的深度知识追踪模型十分具有挑战性.针对此问题,提出一种基于元学习增强的早期知识追踪框架(Meta-Learning-Enhanced Knowledge Tracing,MetaKT).给定目标知识追踪任务和其他若干个相关辅助任务,MetaKT首先在辅助任务上训练模型,然后利用目标任务的数据对预训练后的模型进行微调直至模型收敛.在七个公开数据集上以常用的DKT和DKVMN为基准进行实验,结果发现,提出的MetaKT框架使DKT和DKVMN模型分别在27和33(共35)个测试场景中的AUC(Area under Curve)获得了提升.展开更多
基金This research work is supported by Sichuan Science and Technology Program(Grant No.2022YFS0586)the National Key R&D Program of China(Grant No.2019YFC1509301)the National Natural Science Foundation of China(Grant No.61976046).
文摘Predicting the displacement of landslide is of utmost practical importance as the landslide can pose serious threats to both human life and property.However,traditional methods have the limitation of random selection in sliding window selection and seldom incorporate weather forecast data for displacement prediction,while a single structural model cannot handle input sequences of different lengths at the same time.In order to solve these limitations,in this study,a new approach is proposed that utilizes weather forecast data and incorporates the maximum information coefficient(MIC),long short-term memory network(LSTM),and attention mechanism to establish a teacher-student coupling model with parallel structure for short-term landslide displacement prediction.Through MIC,a suitable input sequence length is selected for the LSTM model.To investigate the influence of rainfall on landslides during different seasons,a parallel teacher-student coupling model is developed that is able to learn sequential information from various time series of different lengths.The teacher model learns sequence information from rainfall intensity time series while incorporating reliable short-term weather forecast data from platforms such as China Meteorological Administration(CMA)and Reliable Prognosis(https://rp5.ru)to improve the model’s expression capability,and the student model learns sequence information from other time series.An attention module is then designed to integrate different sequence information to derive a context vector,representing seasonal temporal attention mode.Finally,the predicted displacement is obtained through a linear layer.The proposed method demonstrates superior prediction accuracies,surpassing those of the support vector machine(SVM),LSTM,recurrent neural network(RNN),temporal convolutional network(TCN),and LSTM-Attention models.It achieves a mean absolute error(MAE)of 0.072 mm,root mean square error(RMSE)of 0.096 mm,and pearson correlation coefficients(PCCS)of 0.85.Additionally,it exhibits enhanced prediction stability and interpretability,rendering it an indispensable tool for landslide disaster prevention and mitigation.
文摘This paper presents the use of a student model to improve the explanations provided by an intelligent tu- toring system,namely SimpleQuestl,in the domain of electronics.The method of overlay modelling is adopted to build the student model.The diagnosis is based on the comparison of the behavinurs of the student and the ex- pert.The student model is consulted by the “explainer” and “debugging” procedures in order to re-order the sequence of the explanation.
文摘It focuses on that students must be developed the ability to solve the practical problem by building the mathematics models and the ability to combine the theory with the practice. It also states that students must be improved the learning interests and practical experience.
文摘In this study, the mathematical models of dynamics of student populations in the university departments are formulated. As a case study, we employ the data of registration section from Department of Mathematics, Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok (KMUTNB), Thailand, from calendar year 2006 to 2010. Using regression analysis, descriptive model and explanatory model are derived. The descriptive model is linear with R2 = 0.8864. Using log-transformation, the explanatory model gives the nonlinear approximation with R2 = 0.8293. The model predicts that the number of students of Department of Mathematics, KMUTNB has a tendency to linearly increase with slope of 20 with 95% CI (6.8417, 33.1583). The application of the models in educational management is discussed.
文摘针对基于知识蒸馏的工业图像异常检测中,预训练网络在迁移到工业图像领域时因领域差异和数据分布的不同而导致获取的特征存在偏差问题,提出了一种特征自适应师生模型。该模型利用特征自适应器来调整教师网络提取的预训练特征,实现跨域特征转换和减少领域偏差。为了避免特征自适应器过度调整预训练特征,导致与原始预训练特征差异过大而降低泛化性能,提出监督损失来约束调整后的特征。此外,进一步为了提高模型对异常特征的表征和判别能力,设计了一个新的学生网络和提出了一种对抗损失来拉远教师和学生网络中异常特征之间的差异,拉近两者正常特征之间的差异。在多个工业数据集MVTec AD、BTAD、VisA和MVTec 3D AD上验证了该模型的有效性。
文摘知识追踪根据学生的历史学习轨迹来实现对其知识掌握水平的实时评估与跟踪,从而预测学生未来的学习表现,是在线学习系统实现个性化学习体验的核心技术之一.与传统方法相比,现有的基于深度神经网络的知识追踪模型取得了显著优势,但其通常依赖大量训练数据.在学生答题早期,交互数据极度稀缺,所以训练一个复杂、有效的深度知识追踪模型十分具有挑战性.针对此问题,提出一种基于元学习增强的早期知识追踪框架(Meta-Learning-Enhanced Knowledge Tracing,MetaKT).给定目标知识追踪任务和其他若干个相关辅助任务,MetaKT首先在辅助任务上训练模型,然后利用目标任务的数据对预训练后的模型进行微调直至模型收敛.在七个公开数据集上以常用的DKT和DKVMN为基准进行实验,结果发现,提出的MetaKT框架使DKT和DKVMN模型分别在27和33(共35)个测试场景中的AUC(Area under Curve)获得了提升.
