Similar to many fields of sciences,recent deep learning advances have been applied extensively in geosciences for both small-and large-scale problems.However,the necessity of using large training data and the’black ...Similar to many fields of sciences,recent deep learning advances have been applied extensively in geosciences for both small-and large-scale problems.However,the necessity of using large training data and the’black box’nature of learning have limited them in practice and difficult to interpret.Furthermore,including the governing equations and physical facts in such methods is also another challenge,which entails either ignoring the physics or simplifying them using unrealistic data.To address such issues,physics informed machine learning methods have been developed which can integrate the governing physics law into the learning process.In this work,a 1-dimensional(1 D)time-dependent seismic wave equation is considered and solved using two methods,namely Gaussian process(GP)and physics informed neural networks.We show that these meshless methods are trained by smaller amount of data and can predict the solution of the equation with even high accuracy.They are also capable of inverting any parameter involved in the governing equation such as wave velocity in our case.Results show that the GP can predict the solution of the seismic wave equation with a lower level of error,while our developed neural network is more accurate for velocity(P-and S-wave)and density inversion.展开更多
This paper investigates the dynamic evolution with limited learning information on a small-world network.In the system, the information among the interaction players is not very lucid, and the players are not allowed ...This paper investigates the dynamic evolution with limited learning information on a small-world network.In the system, the information among the interaction players is not very lucid, and the players are not allowed to inspectthe profit collected by its neighbors, thus the focal player cannot choose randomly a neighbor or the wealthiest one andcompare its payoff to copy its strategy.It is assumed that the information acquainted by the player declines in theform of the exponential with the geographical distance between the players, and a parameter V is introduced to denotethe inspect-ability about the players.It is found that under the hospitable conditions, cooperation increases with therandomness and is inhibited by the large connectivity for the prisoner's dilemma; however, cooperation is maximal atthe moderate rewiring probability and is chaos with the connectivity for the snowdrift game.For the two games, theacuminous sight is in favor of the cooperation under the hospitable conditions; whereas, the myopic eyes are advantageousto cooperation and cooperation increases with the randomness under the hostile condition.展开更多
Mobile-learning can help learners select favorite learning resources at anytime and anywhere,and breaks the limitation of time and space and increase students'interests and initiatives.But it also causes some prob...Mobile-learning can help learners select favorite learning resources at anytime and anywhere,and breaks the limitation of time and space and increase students'interests and initiatives.But it also causes some problems.How to use M-learning to improve the efficiency of vocabulary learning has drawn much attention.Based on theoretical discussion,this paper carried out a survey on university students'vocabulary M-learning situation,the effectiveness and their attitude to M-learning.And then it puts forward some suggestions about vocabulary learning.展开更多
Starting from presenting and analyzing some information gap activities during the previous teaching experience, this article has inferred the major roles of information gap activities. Some strategies to implement the...Starting from presenting and analyzing some information gap activities during the previous teaching experience, this article has inferred the major roles of information gap activities. Some strategies to implement the information gap activities are also recommended together with the functions of the instructors via these activities. What information gap activities can teach us in TESOL （teaching English for speakers of other languages） is that information gap activities contribute to setting up a climate of a mutual autonomous learning style both for the learners and the instructors, and these activities activate a diversity in the learning atmosphere.展开更多
Geology is perhaps the most fascinating of the natural sciences, due to its all-encompassing nature. Virtually all human activities that occur on planet Earth--including agriculture, energy and mineral resource explor...Geology is perhaps the most fascinating of the natural sciences, due to its all-encompassing nature. Virtually all human activities that occur on planet Earth--including agriculture, energy and mineral resource exploration and extraction; environmental and public policy on natural resources management and protection; land use planning; infrastructure development; and ecological tourism--all depend on various aspects of geology and its sub-disciplines. Due to the importance of geology in the daily lives of human beings, it is imperative that all persons develop at least a basic understanding of the science. In this paper, the current efforts for promoting public understanding in geology will be examined, with offerings of alternatives and supplements to these efforts. Information from the science education sub-disciplines of HPS （history, philosophy and sociology） of science, and informal/free-choice learning will be woven into the framework of the geology-public understanding idea.展开更多
Based on the concept of learning stickiness,this study constructed a model of influencing factors among the elements of online learning content,interaction,incentive,satisfaction,and learning stickiness from Comprehen...Based on the concept of learning stickiness,this study constructed a model of influencing factors among the elements of online learning content,interaction,incentive,satisfaction,and learning stickiness from Comprehensive Learning Theory.Structural equation modeling was used to analyze the interaction and influence effects among the factors.It is found that the content,interaction,and incentive in Comprehensive Learning Theory had a significant positive impact on learning stickiness from the total effect analysis.From the direct effect analysis,the influence of content and interaction on learning stickiness was not substantial,but the influence of incentive and satisfaction on learning stickiness was significant.From the perspective of mediation effect analysis,incentive and satisfaction were critical mediating variables for the influence of content and interaction on learning stickiness.This study put forward suggestions and strategies for online teaching,providing a reference for teachers to carry out online education.展开更多
A novel neural network method is developed for solving systems of conservation laws whose solutions may contain abrupt changes of state,including shock waves and contact discontinuities.In conventional approaches,a lo...A novel neural network method is developed for solving systems of conservation laws whose solutions may contain abrupt changes of state,including shock waves and contact discontinuities.In conventional approaches,a low-cost solution patch is usually used as the input to a neural network for predicting the high-fidelity solution patch.With that technique,however,there is no way to distinguish a smeared discontinuity from a smooth solution with large gradient in the input,and the two almost identical inputs correspond to two fundamentally different high-fidelity solution patches in training and predicting.To circumvent this difficulty,we use local patches of two low-cost numerical solutions of the conservation laws in a converging sequence as the input to a neural network.The neural network then makes a correct prediction by identifying whether the solution contains discontinuities or just smooth variations with large gradients,because the former becomes increasingly steep in a converging sequence in the input,and the latter does not.The inputs can be computed from lowcost numerical schemes with coarse resolution,in a local domain of dependence of a space-time location where the prediction is to be made.Despite smeared input solutions,the output provides sharp approximations of solutions containing shock waves and contact discontinuities.The method works effectively not only for regions with discontinuities,but also for smooth regions of the solution.It is efficient to implement,once trained,and has broader applications for different types of differential equations.展开更多
In our prior work[10],neural networks with local converging inputs(NNLCI)were introduced for solving one-dimensional conservation equations.Two solutions of a conservation law in a converging sequence,computed from lo...In our prior work[10],neural networks with local converging inputs(NNLCI)were introduced for solving one-dimensional conservation equations.Two solutions of a conservation law in a converging sequence,computed from low-cost numerical schemes,and in a local domain of dependence of the space-time location,were used as the input to a neural network in order to predict a high-fidelity solution at a given space-time location.In the present work,we extend the method to twodimensional conservation systems and introduce different solution techniques.Numerical results demonstrate the validity and effectiveness of the NNLCI method for application to multi-dimensional problems.In spite of low-cost smeared input data,the NNLCI method is capable of accurately predicting shocks,contact discontinuities,and the smooth region of the entire field.The NNLCI method is relatively easy to train because of the use of local solvers.The computing time saving is between one and two orders of magnitude compared with the corresponding high-fidelity schemes for two-dimensional Riemann problems.The relative efficiency of the NNLCI method is expected to be substantially greater for problems with higher spatial dimensions or smooth solutions.展开更多
In recent years,China has been issuing artificial intelligence(AI)education policies to promote the deep integration between AI and education.Setting AI courses at the basic education phase can cultivate students’AI ...In recent years,China has been issuing artificial intelligence(AI)education policies to promote the deep integration between AI and education.Setting AI courses at the basic education phase can cultivate students’AI literacy and enhance the country’s science and technology competitiveness in the future.However,due to shortages of AI teachers and inadequate conditions for practice,the current AI courses in primary and secondary schools are not enough in effectiveness.In light of current problems,this study,based on the foundation of university-enterprise cooperation,integrates the“competition and education”mechanism into informal AI course design and practice.Using a design-based research paradigm,the study proposes the design framework from four dimensions:goals,themes and content,methods and steps,and evaluation,and then further refines the framework through two rounds of iterations,to provide a reference for the development and practice of AI courses.展开更多
Neural networks have been widely used for English name tagging and have delivered state-of-the-art results. However, for low resource languages, due to the limited resources and lack of training data, taggers tend to ...Neural networks have been widely used for English name tagging and have delivered state-of-the-art results. However, for low resource languages, due to the limited resources and lack of training data, taggers tend to have lower performance, in comparison to the English language. In this paper, we tackle this challenging issue by incorporating multi-level cross-lingual knowledge as attention into a neural architecture, which guides low resource name tagging to achieve a better performance. Specifically, we regard entity type distribution as language independent and use bilingual lexicons to bridge cross-lingual semantic mapping. Then, we jointly apply word-level cross-lingual mutual influence and entity-type level monolingual word distributions to enhance low resource name tagging. Experiments on three languages demonstrate the effectiveness of this neural architecture： for Chinese,Uzbek, and Turkish, we are able to yield significant improvements in name tagging over all previous baselines.展开更多
文摘Similar to many fields of sciences,recent deep learning advances have been applied extensively in geosciences for both small-and large-scale problems.However,the necessity of using large training data and the’black box’nature of learning have limited them in practice and difficult to interpret.Furthermore,including the governing equations and physical facts in such methods is also another challenge,which entails either ignoring the physics or simplifying them using unrealistic data.To address such issues,physics informed machine learning methods have been developed which can integrate the governing physics law into the learning process.In this work,a 1-dimensional(1 D)time-dependent seismic wave equation is considered and solved using two methods,namely Gaussian process(GP)and physics informed neural networks.We show that these meshless methods are trained by smaller amount of data and can predict the solution of the equation with even high accuracy.They are also capable of inverting any parameter involved in the governing equation such as wave velocity in our case.Results show that the GP can predict the solution of the seismic wave equation with a lower level of error,while our developed neural network is more accurate for velocity(P-and S-wave)and density inversion.
基金Supported by Natural Science Foundation of China under Grant No.10974146
文摘This paper investigates the dynamic evolution with limited learning information on a small-world network.In the system, the information among the interaction players is not very lucid, and the players are not allowed to inspectthe profit collected by its neighbors, thus the focal player cannot choose randomly a neighbor or the wealthiest one andcompare its payoff to copy its strategy.It is assumed that the information acquainted by the player declines in theform of the exponential with the geographical distance between the players, and a parameter V is introduced to denotethe inspect-ability about the players.It is found that under the hospitable conditions, cooperation increases with therandomness and is inhibited by the large connectivity for the prisoner's dilemma; however, cooperation is maximal atthe moderate rewiring probability and is chaos with the connectivity for the snowdrift game.For the two games, theacuminous sight is in favor of the cooperation under the hospitable conditions; whereas, the myopic eyes are advantageousto cooperation and cooperation increases with the randomness under the hostile condition.
文摘Mobile-learning can help learners select favorite learning resources at anytime and anywhere,and breaks the limitation of time and space and increase students'interests and initiatives.But it also causes some problems.How to use M-learning to improve the efficiency of vocabulary learning has drawn much attention.Based on theoretical discussion,this paper carried out a survey on university students'vocabulary M-learning situation,the effectiveness and their attitude to M-learning.And then it puts forward some suggestions about vocabulary learning.
文摘Starting from presenting and analyzing some information gap activities during the previous teaching experience, this article has inferred the major roles of information gap activities. Some strategies to implement the information gap activities are also recommended together with the functions of the instructors via these activities. What information gap activities can teach us in TESOL （teaching English for speakers of other languages） is that information gap activities contribute to setting up a climate of a mutual autonomous learning style both for the learners and the instructors, and these activities activate a diversity in the learning atmosphere.
文摘Geology is perhaps the most fascinating of the natural sciences, due to its all-encompassing nature. Virtually all human activities that occur on planet Earth--including agriculture, energy and mineral resource exploration and extraction; environmental and public policy on natural resources management and protection; land use planning; infrastructure development; and ecological tourism--all depend on various aspects of geology and its sub-disciplines. Due to the importance of geology in the daily lives of human beings, it is imperative that all persons develop at least a basic understanding of the science. In this paper, the current efforts for promoting public understanding in geology will be examined, with offerings of alternatives and supplements to these efforts. Information from the science education sub-disciplines of HPS （history, philosophy and sociology） of science, and informal/free-choice learning will be woven into the framework of the geology-public understanding idea.
文摘Based on the concept of learning stickiness,this study constructed a model of influencing factors among the elements of online learning content,interaction,incentive,satisfaction,and learning stickiness from Comprehensive Learning Theory.Structural equation modeling was used to analyze the interaction and influence effects among the factors.It is found that the content,interaction,and incentive in Comprehensive Learning Theory had a significant positive impact on learning stickiness from the total effect analysis.From the direct effect analysis,the influence of content and interaction on learning stickiness was not substantial,but the influence of incentive and satisfaction on learning stickiness was significant.From the perspective of mediation effect analysis,incentive and satisfaction were critical mediating variables for the influence of content and interaction on learning stickiness.This study put forward suggestions and strategies for online teaching,providing a reference for teachers to carry out online education.
基金supported in part by NSF grant DMS-1522585partly sponsored by the Ralph N.Read Endowment of the Georgia Institute of Technology.
文摘A novel neural network method is developed for solving systems of conservation laws whose solutions may contain abrupt changes of state,including shock waves and contact discontinuities.In conventional approaches,a low-cost solution patch is usually used as the input to a neural network for predicting the high-fidelity solution patch.With that technique,however,there is no way to distinguish a smeared discontinuity from a smooth solution with large gradient in the input,and the two almost identical inputs correspond to two fundamentally different high-fidelity solution patches in training and predicting.To circumvent this difficulty,we use local patches of two low-cost numerical solutions of the conservation laws in a converging sequence as the input to a neural network.The neural network then makes a correct prediction by identifying whether the solution contains discontinuities or just smooth variations with large gradients,because the former becomes increasingly steep in a converging sequence in the input,and the latter does not.The inputs can be computed from lowcost numerical schemes with coarse resolution,in a local domain of dependence of a space-time location where the prediction is to be made.Despite smeared input solutions,the output provides sharp approximations of solutions containing shock waves and contact discontinuities.The method works effectively not only for regions with discontinuities,but also for smooth regions of the solution.It is efficient to implement,once trained,and has broader applications for different types of differential equations.
文摘In our prior work[10],neural networks with local converging inputs(NNLCI)were introduced for solving one-dimensional conservation equations.Two solutions of a conservation law in a converging sequence,computed from low-cost numerical schemes,and in a local domain of dependence of the space-time location,were used as the input to a neural network in order to predict a high-fidelity solution at a given space-time location.In the present work,we extend the method to twodimensional conservation systems and introduce different solution techniques.Numerical results demonstrate the validity and effectiveness of the NNLCI method for application to multi-dimensional problems.In spite of low-cost smeared input data,the NNLCI method is capable of accurately predicting shocks,contact discontinuities,and the smooth region of the entire field.The NNLCI method is relatively easy to train because of the use of local solvers.The computing time saving is between one and two orders of magnitude compared with the corresponding high-fidelity schemes for two-dimensional Riemann problems.The relative efficiency of the NNLCI method is expected to be substantially greater for problems with higher spatial dimensions or smooth solutions.
文摘In recent years,China has been issuing artificial intelligence(AI)education policies to promote the deep integration between AI and education.Setting AI courses at the basic education phase can cultivate students’AI literacy and enhance the country’s science and technology competitiveness in the future.However,due to shortages of AI teachers and inadequate conditions for practice,the current AI courses in primary and secondary schools are not enough in effectiveness.In light of current problems,this study,based on the foundation of university-enterprise cooperation,integrates the“competition and education”mechanism into informal AI course design and practice.Using a design-based research paradigm,the study proposes the design framework from four dimensions:goals,themes and content,methods and steps,and evaluation,and then further refines the framework through two rounds of iterations,to provide a reference for the development and practice of AI courses.
基金supported by the National High-Tech Development(863)Program of China(No.2015AA015407)the National Natural Science Foundation of China(Nos.61632011 and 61370164)
文摘Neural networks have been widely used for English name tagging and have delivered state-of-the-art results. However, for low resource languages, due to the limited resources and lack of training data, taggers tend to have lower performance, in comparison to the English language. In this paper, we tackle this challenging issue by incorporating multi-level cross-lingual knowledge as attention into a neural architecture, which guides low resource name tagging to achieve a better performance. Specifically, we regard entity type distribution as language independent and use bilingual lexicons to bridge cross-lingual semantic mapping. Then, we jointly apply word-level cross-lingual mutual influence and entity-type level monolingual word distributions to enhance low resource name tagging. Experiments on three languages demonstrate the effectiveness of this neural architecture： for Chinese,Uzbek, and Turkish, we are able to yield significant improvements in name tagging over all previous baselines.
