Traditional expert-designed branching rules in branch-and-bound(B&B) are static, often failing to adapt to diverse and evolving problem instances. Crafting these rules is labor-intensive, and may not scale well wi...Traditional expert-designed branching rules in branch-and-bound(B&B) are static, often failing to adapt to diverse and evolving problem instances. Crafting these rules is labor-intensive, and may not scale well with complex problems.Given the frequent need to solve varied combinatorial optimization problems, leveraging statistical learning to auto-tune B&B algorithms for specific problem classes becomes attractive. This paper proposes a graph pointer network model to learn the branch rules. Graph features, global features and historical features are designated to represent the solver state. The graph neural network processes graph features, while the pointer mechanism assimilates the global and historical features to finally determine the variable on which to branch. The model is trained to imitate the expert strong branching rule by a tailored top-k Kullback-Leibler divergence loss function. Experiments on a series of benchmark problems demonstrate that the proposed approach significantly outperforms the widely used expert-designed branching rules. It also outperforms state-of-the-art machine-learning-based branch-and-bound methods in terms of solving speed and search tree size on all the test instances. In addition, the model can generalize to unseen instances and scale to larger instances.展开更多
The theory of“imitation”in painting occupies a leading position in western art,which originated from the theory of“imitation”in ancient Greece,and has become one of the art theories affecting the world through the...The theory of“imitation”in painting occupies a leading position in western art,which originated from the theory of“imitation”in ancient Greece,and has become one of the art theories affecting the world through the continuous development of later generations.Through the exploration of the source of“imitation”in China and the West,there are some comments on the meaning of“imitation”in Chinese classical painting theory,such as“transfer model writing”and“image form”,which is obvious differences from the west.Traditional Chinese painting is a combination of careful observation of natural things and subjective emotions to express their own aesthetic feelings,and ultimately form a vivid artistic conception.Modern imitation is borrowed from Western imitation.In fact,imitation in traditional painting has its own meaning,which contains Chinese aesthetic thought.“Imitation”aesthetics is unique in traditional Chinese painting and is the most important form of painting art.展开更多
The Saghro massif constitutes a vast metallogenic province with numerous deposits and shows of base metals (lead, zinc, copper) and precious metals (gold and silver), besides various useful substances (talc, pyrophyll...The Saghro massif constitutes a vast metallogenic province with numerous deposits and shows of base metals (lead, zinc, copper) and precious metals (gold and silver), besides various useful substances (talc, pyrophyllite, barite, fluorite). Silver/lead occurrences are concentrated along the Cryogenian Imiter series and moderately at Boumalne and Sidi Flah. Copper occupies the plutonic intrusions and intrusive rocks of the East-Central Saghro while barite deposits are widespread throughout the Cambrian cover of the East Saghro in contact with the Ediacaran basement. To justify this distribution, the new contributions of the cartography and the organic geochemistry of the black shales of Jbel Saghro have clearly shown the particularity of the Imiter black shales in terms of the richness in organic matter (TOC = 0.18%), the blackish color and the friability. The Boumalne and Sidi Flah groups present some similarities with the Imiter group, such as the sub-equatorial structuring, the friable pelites and the richness in organic matter (Boumalne TOC = 0.11% and SidiFlah TOC = 0.16%), which is a quite good show that requires to reinforce the exploration works. For Western Saghro in the Iknioun and Qalaa’t M’Gouna groups, the variations in the thickness of the volcanic cover show an irregular paleotopography with hard, greenish, organic-poor pelitic sediments (TOC = 0.01 to 0.04%). We can conclude that the formation of Imiter-type silver concentrations requires the combination of the sedimentological, the volcanic and structural factors. For Imiter-type silver these factors are: a fine pelitic and argillic casing deposited in a confined environment, a basic volcanism source of metals and other intermediate to acid generated by the hydrothermalism and heat, a convenable paleotopography and a network of fracturations to trap the mineralizations.展开更多
Memetics confirms the importance of recitation and imitation in college English teaching.Starting with the explanation of the replication cycle of language memes,this paper discusses how to use the memetic perspective...Memetics confirms the importance of recitation and imitation in college English teaching.Starting with the explanation of the replication cycle of language memes,this paper discusses how to use the memetic perspective in college English listening and speaking teaching,effectively transform the input language information into language output,and improve the classroom effect of college English listening and speaking teaching.展开更多
Aim The particle texture from diesel engine was imitated by use of computer. Methods The theory of fractal geometry and the diffusion limited aggregation model were used to simulate the micron texture. Results The...Aim The particle texture from diesel engine was imitated by use of computer. Methods The theory of fractal geometry and the diffusion limited aggregation model were used to simulate the micron texture. Results The fractal dimensions of granule distribution and corpuscle superficial area are quite conformed with those of measurement. Conclusion The texture parameters of engine particle cluster can be obtained precisely by use of fractal theory.展开更多
UG and imitation are two parallel hypotheses trying to answer how childrens language acquisition is realized. Imitation fails to explain how children acquire language; however, it helps a lot in childrens language acq...UG and imitation are two parallel hypotheses trying to answer how childrens language acquisition is realized. Imitation fails to explain how children acquire language; however, it helps a lot in childrens language acquisition.展开更多
It is of vital importance for modern college English teaching to properly construct an interactive multimedia-internet-based teaching system, the structure of which is clearly elaborated in this paper. An IMITS usuall...It is of vital importance for modern college English teaching to properly construct an interactive multimedia-internet-based teaching system, the structure of which is clearly elaborated in this paper. An IMITS usually consists of hardware, software, teaching and management. At the end of this paper, a conclusion is made that only when all the four parts of IMITS are construct ed such as is demonstrated, can the IMITS exert its full effects in college English teaching.展开更多
In this study,a 3D virtual reality and visualization engine for rendering the ocean,named VV-Ocean,is designed for marine applications.The design goals of VV-Ocean aim at high fidelity simulation of ocean environment,...In this study,a 3D virtual reality and visualization engine for rendering the ocean,named VV-Ocean,is designed for marine applications.The design goals of VV-Ocean aim at high fidelity simulation of ocean environment,visualization of massive and multidimensional marine data,and imitation of marine lives.VV-Ocean is composed of five modules,i.e.memory management module,resources management module,scene management module,rendering process management module and interaction management module.There are three core functions in VV-Ocean:reconstructing vivid virtual ocean scenes,visualizing real data dynamically in real time,imitating and simulating marine lives intuitively.Based on VV-Ocean,we establish a sea-land integration platform which can reproduce drifting and diffusion processes of oil spilling from sea bottom to surface.Environment factors such as ocean current and wind field have been considered in this simulation.On this platform oil spilling process can be abstracted as movements of abundant oil particles.The result shows that oil particles blend with water well and the platform meets the requirement for real-time and interactive rendering.VV-Ocean can be widely used in ocean applications such as demonstrating marine operations,facilitating maritime communications,developing ocean games,reducing marine hazards,forecasting the weather over oceans,serving marine tourism,and so on.Finally,further technological improvements of VV-Ocean are discussed.展开更多
One of the assumptions of previous research in evolutionary game dynamics is that individuals use only one rule to update their strategy. In reality, an individual's strategy update rules may change with the envir...One of the assumptions of previous research in evolutionary game dynamics is that individuals use only one rule to update their strategy. In reality, an individual's strategy update rules may change with the environment, and it is possible for an individual to use two or more rules to update their strategy. We consider the case where an individual updates strategies based on the Moran and imitation processes, and establish mixed stochastic evolutionary game dynamics by combining both processes. Our aim is to study how individuals change strategies based on two update rules and how this affects evolutionary game dynamics. We obtain an analytic expression and properties of the fixation probability and fixation times(the unconditional fixation time or conditional average fixation time) associated with our proposed process. We find unexpected results. The fixation probability within the proposed model is independent of the probabilities that the individual adopts the imitation rule update strategy. This implies that the fixation probability within the proposed model is equal to that from the Moran and imitation processes. The one-third rule holds in the proposed mixed model. However, under weak selection, the fixation times are different from those of the Moran and imitation processes because it is connected with the probability that individuals adopt an imitation update rule. Numerical examples are presented to illustrate the relationships between fixation times and the probability that an individual adopts the imitation update rule, as well as between fixation times and selection intensity. From the simulated analysis, we find that the fixation time for a mixed process is greater than that of the Moran process, but is less than that of the imitation process. Moreover, the fixation times for a cooperator in the proposed process increase as the probability of adopting an imitation update increases; however, the relationship becomes more complex than a linear relationship.展开更多
Mobile Edge Computing(MEC)is promising to alleviate the computation and storage burdens for terminals in wireless networks.The huge energy consumption of MEC servers challenges the establishment of smart cities and th...Mobile Edge Computing(MEC)is promising to alleviate the computation and storage burdens for terminals in wireless networks.The huge energy consumption of MEC servers challenges the establishment of smart cities and their service time powered by rechargeable batteries.In addition,Orthogonal Multiple Access(OMA)technique cannot utilize limited spectrum resources fully and efficiently.Therefore,Non-Orthogonal Multiple Access(NOMA)-based energy-efficient task scheduling among MEC servers for delay-constraint mobile applications is important,especially in highly-dynamic vehicular edge computing networks.The various movement patterns of vehicles lead to unbalanced offloading requirements and different load pressure for MEC servers.Self-Imitation Learning(SIL)-based Deep Reinforcement Learning(DRL)has emerged as a promising machine learning technique to break through obstacles in various research fields,especially in time-varying networks.In this paper,we first introduce related MEC technologies in vehicular networks.Then,we propose an energy-efficient approach for task scheduling in vehicular edge computing networks based on DRL,with the purpose of both guaranteeing the task latency requirement for multiple users and minimizing total energy consumption of MEC servers.Numerical results demonstrate that the proposed algorithm outperforms other methods.展开更多
The zone of proximal development(ZPD) and the scaffolding theory are very different,both in terms of their theoretical origins and connotations,and can even be said to be very different.However,during the development ...The zone of proximal development(ZPD) and the scaffolding theory are very different,both in terms of their theoretical origins and connotations,and can even be said to be very different.However,during the development of the two concepts,some scholars have misunderstood them,resulting in the two being mistaken for similar concepts and therefore often confused.Professor James Lantolf from Pennsylvania State University(State College,USA) was interviewed by Professor Lili Qin from Dalian University of Foreign Studies(Dalian,China) and provides an indepth analysis of these issues.The interview begins with the theoretical roots,connotations and definitions of the ZPD and scaffolding concepts,and then unravels the story of how they have been“mistakenly loved for life”,and ultimately it is made clear that the two concepts are completely different in their practical application to language teaching and should not continue to be used interchangeably.展开更多
Here,the challenges of sample efficiency and navigation performance in deep rein-forcement learning for visual navigation are focused and a deep imitation reinforcement learning approach is proposed.Our contributions ...Here,the challenges of sample efficiency and navigation performance in deep rein-forcement learning for visual navigation are focused and a deep imitation reinforcement learning approach is proposed.Our contributions are mainly three folds:first,a frame-work combining imitation learning with deep reinforcement learning is presented,which enables a robot to learn a stable navigation policy faster in the target-driven navigation task.Second,the surrounding images is taken as the observation instead of sequential images,which can improve the navigation performance for more information.Moreover,a simple yet efficient template matching method is adopted to determine the stop action,making the system more practical.Simulation experiments in the AI-THOR environment show that the proposed approach outperforms previous end-to-end deep reinforcement learning approaches,which demonstrate the effectiveness and efficiency of our approach.展开更多
Providing autonomous systems with an effective quantity and quality of information from a desired task is challenging. In particular, autonomous vehicles, must have a reliable vision of their workspace to robustly acc...Providing autonomous systems with an effective quantity and quality of information from a desired task is challenging. In particular, autonomous vehicles, must have a reliable vision of their workspace to robustly accomplish driving functions. Speaking of machine vision, deep learning techniques, and specifically convolutional neural networks, have been proven to be the state of the art technology in the field. As these networks typically involve millions of parameters and elements, designing an optimal architecture for deep learning structures is a difficult task which is globally under investigation by researchers. This study experimentally evaluates the impact of three major architectural properties of convolutional networks, including the number of layers, filters, and filter size on their performance. In this study, several models with different properties are developed,equally trained, and then applied to an autonomous car in a realistic simulation environment. A new ensemble approach is also proposed to calculate and update weights for the models regarding their mean squared error values. Based on design properties,performance results are reported and compared for further investigations. Surprisingly, the number of filters itself does not largely affect the performance efficiency. As a result, proper allocation of filters with different kernel sizes through the layers introduces a considerable improvement in the performance.Achievements of this study will provide the researchers with a clear clue and direction in designing optimal network architectures for deep learning purposes.展开更多
This paper reports on a study on the effects of reading-writing integrated tasks on vocabulary learning and explored the differential roles of creative construction and non-creative construction in promoting lexical l...This paper reports on a study on the effects of reading-writing integrated tasks on vocabulary learning and explored the differential roles of creative construction and non-creative construction in promoting lexical learning. Participants were 90 first-year English majors, randomly assigned to two experimental groups(continuation and retelling) and one control group, with 30 students in each group. Results showed that the continuation group generated a substantial amount of creative construction and produced significantly more instances of creative imitation than the retelling group. The continuation group outperformed the retelling group for both receptive and productive vocabulary knowledge gain and retention, but differences were only significant in terms of productive vocabulary retention. Finally, productive vocabulary knowledge retention among the continuation group was significantly and positively correlated with creative imitation(meaning creation coupled with language imitation), but not with linguistic alignment per se. As productive vocabulary knowledge constitutes the learner ’s ability to use lexical knowledge to express ideas in dynamic contexts, the findings afforded evidence that creative imitation could be the answer to the fundamental issue of L2 learning(i.e., mapping static language onto dynamic idea expression). The pedagogical implications as well as future research directions are also discussed.展开更多
Night,Mother is a famous play about female identification.It is about Jessie,a middle aged woman,who,through many years’struggle,decides to kill herself at nine o’clock in an ordinary night in her mother’s house.Je...Night,Mother is a famous play about female identification.It is about Jessie,a middle aged woman,who,through many years’struggle,decides to kill herself at nine o’clock in an ordinary night in her mother’s house.Jessie’s fighting for female identity and corresponding power would be analyzed.In the play,Jessie tries two ways to get her power,one by trying to imitating men to get accepted by patriarchal society,and the other by revolting outside resistance and insisting her pursuit of self rights and freedom.展开更多
基金supported by the Open Project of Xiangjiang Laboratory (22XJ02003)Scientific Project of the National University of Defense Technology (NUDT)(ZK21-07, 23-ZZCX-JDZ-28)+1 种基金the National Science Fund for Outstanding Young Scholars (62122093)the National Natural Science Foundation of China (72071205)。
文摘Traditional expert-designed branching rules in branch-and-bound(B&B) are static, often failing to adapt to diverse and evolving problem instances. Crafting these rules is labor-intensive, and may not scale well with complex problems.Given the frequent need to solve varied combinatorial optimization problems, leveraging statistical learning to auto-tune B&B algorithms for specific problem classes becomes attractive. This paper proposes a graph pointer network model to learn the branch rules. Graph features, global features and historical features are designated to represent the solver state. The graph neural network processes graph features, while the pointer mechanism assimilates the global and historical features to finally determine the variable on which to branch. The model is trained to imitate the expert strong branching rule by a tailored top-k Kullback-Leibler divergence loss function. Experiments on a series of benchmark problems demonstrate that the proposed approach significantly outperforms the widely used expert-designed branching rules. It also outperforms state-of-the-art machine-learning-based branch-and-bound methods in terms of solving speed and search tree size on all the test instances. In addition, the model can generalize to unseen instances and scale to larger instances.
文摘The theory of“imitation”in painting occupies a leading position in western art,which originated from the theory of“imitation”in ancient Greece,and has become one of the art theories affecting the world through the continuous development of later generations.Through the exploration of the source of“imitation”in China and the West,there are some comments on the meaning of“imitation”in Chinese classical painting theory,such as“transfer model writing”and“image form”,which is obvious differences from the west.Traditional Chinese painting is a combination of careful observation of natural things and subjective emotions to express their own aesthetic feelings,and ultimately form a vivid artistic conception.Modern imitation is borrowed from Western imitation.In fact,imitation in traditional painting has its own meaning,which contains Chinese aesthetic thought.“Imitation”aesthetics is unique in traditional Chinese painting and is the most important form of painting art.
文摘The Saghro massif constitutes a vast metallogenic province with numerous deposits and shows of base metals (lead, zinc, copper) and precious metals (gold and silver), besides various useful substances (talc, pyrophyllite, barite, fluorite). Silver/lead occurrences are concentrated along the Cryogenian Imiter series and moderately at Boumalne and Sidi Flah. Copper occupies the plutonic intrusions and intrusive rocks of the East-Central Saghro while barite deposits are widespread throughout the Cambrian cover of the East Saghro in contact with the Ediacaran basement. To justify this distribution, the new contributions of the cartography and the organic geochemistry of the black shales of Jbel Saghro have clearly shown the particularity of the Imiter black shales in terms of the richness in organic matter (TOC = 0.18%), the blackish color and the friability. The Boumalne and Sidi Flah groups present some similarities with the Imiter group, such as the sub-equatorial structuring, the friable pelites and the richness in organic matter (Boumalne TOC = 0.11% and SidiFlah TOC = 0.16%), which is a quite good show that requires to reinforce the exploration works. For Western Saghro in the Iknioun and Qalaa’t M’Gouna groups, the variations in the thickness of the volcanic cover show an irregular paleotopography with hard, greenish, organic-poor pelitic sediments (TOC = 0.01 to 0.04%). We can conclude that the formation of Imiter-type silver concentrations requires the combination of the sedimentological, the volcanic and structural factors. For Imiter-type silver these factors are: a fine pelitic and argillic casing deposited in a confined environment, a basic volcanism source of metals and other intermediate to acid generated by the hydrothermalism and heat, a convenable paleotopography and a network of fracturations to trap the mineralizations.
文摘Memetics confirms the importance of recitation and imitation in college English teaching.Starting with the explanation of the replication cycle of language memes,this paper discusses how to use the memetic perspective in college English listening and speaking teaching,effectively transform the input language information into language output,and improve the classroom effect of college English listening and speaking teaching.
文摘Aim The particle texture from diesel engine was imitated by use of computer. Methods The theory of fractal geometry and the diffusion limited aggregation model were used to simulate the micron texture. Results The fractal dimensions of granule distribution and corpuscle superficial area are quite conformed with those of measurement. Conclusion The texture parameters of engine particle cluster can be obtained precisely by use of fractal theory.
文摘UG and imitation are two parallel hypotheses trying to answer how childrens language acquisition is realized. Imitation fails to explain how children acquire language; however, it helps a lot in childrens language acquisition.
文摘It is of vital importance for modern college English teaching to properly construct an interactive multimedia-internet-based teaching system, the structure of which is clearly elaborated in this paper. An IMITS usually consists of hardware, software, teaching and management. At the end of this paper, a conclusion is made that only when all the four parts of IMITS are construct ed such as is demonstrated, can the IMITS exert its full effects in college English teaching.
基金supported by the Global Change Research Program of China under Project 2012CB955603the Natural Science Foundation of China under Project 41076115+2 种基金the National Basic Research Program of China under Project 2009CB723903the Public Science and Technology Research Funds of the Ocean under Project 201005019the National High-Tech Research and Development Program of China under Project 2008AA121701
文摘In this study,a 3D virtual reality and visualization engine for rendering the ocean,named VV-Ocean,is designed for marine applications.The design goals of VV-Ocean aim at high fidelity simulation of ocean environment,visualization of massive and multidimensional marine data,and imitation of marine lives.VV-Ocean is composed of five modules,i.e.memory management module,resources management module,scene management module,rendering process management module and interaction management module.There are three core functions in VV-Ocean:reconstructing vivid virtual ocean scenes,visualizing real data dynamically in real time,imitating and simulating marine lives intuitively.Based on VV-Ocean,we establish a sea-land integration platform which can reproduce drifting and diffusion processes of oil spilling from sea bottom to surface.Environment factors such as ocean current and wind field have been considered in this simulation.On this platform oil spilling process can be abstracted as movements of abundant oil particles.The result shows that oil particles blend with water well and the platform meets the requirement for real-time and interactive rendering.VV-Ocean can be widely used in ocean applications such as demonstrating marine operations,facilitating maritime communications,developing ocean games,reducing marine hazards,forecasting the weather over oceans,serving marine tourism,and so on.Finally,further technological improvements of VV-Ocean are discussed.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.71871171,71871173,and 71832010)
文摘One of the assumptions of previous research in evolutionary game dynamics is that individuals use only one rule to update their strategy. In reality, an individual's strategy update rules may change with the environment, and it is possible for an individual to use two or more rules to update their strategy. We consider the case where an individual updates strategies based on the Moran and imitation processes, and establish mixed stochastic evolutionary game dynamics by combining both processes. Our aim is to study how individuals change strategies based on two update rules and how this affects evolutionary game dynamics. We obtain an analytic expression and properties of the fixation probability and fixation times(the unconditional fixation time or conditional average fixation time) associated with our proposed process. We find unexpected results. The fixation probability within the proposed model is independent of the probabilities that the individual adopts the imitation rule update strategy. This implies that the fixation probability within the proposed model is equal to that from the Moran and imitation processes. The one-third rule holds in the proposed mixed model. However, under weak selection, the fixation times are different from those of the Moran and imitation processes because it is connected with the probability that individuals adopt an imitation update rule. Numerical examples are presented to illustrate the relationships between fixation times and the probability that an individual adopts the imitation update rule, as well as between fixation times and selection intensity. From the simulated analysis, we find that the fixation time for a mixed process is greater than that of the Moran process, but is less than that of the imitation process. Moreover, the fixation times for a cooperator in the proposed process increase as the probability of adopting an imitation update increases; however, the relationship becomes more complex than a linear relationship.
基金supported in part by the National Natural Science Foundation of China under Grant 61971084 and Grant 62001073in part by the National Natural Science Foundation of Chongqing under Grant cstc2019jcyj-msxmX0208in part by the open research fund of National Mobile Communications Research Laboratory,Southeast University,under Grant 2020D05.
文摘Mobile Edge Computing(MEC)is promising to alleviate the computation and storage burdens for terminals in wireless networks.The huge energy consumption of MEC servers challenges the establishment of smart cities and their service time powered by rechargeable batteries.In addition,Orthogonal Multiple Access(OMA)technique cannot utilize limited spectrum resources fully and efficiently.Therefore,Non-Orthogonal Multiple Access(NOMA)-based energy-efficient task scheduling among MEC servers for delay-constraint mobile applications is important,especially in highly-dynamic vehicular edge computing networks.The various movement patterns of vehicles lead to unbalanced offloading requirements and different load pressure for MEC servers.Self-Imitation Learning(SIL)-based Deep Reinforcement Learning(DRL)has emerged as a promising machine learning technique to break through obstacles in various research fields,especially in time-varying networks.In this paper,we first introduce related MEC technologies in vehicular networks.Then,we propose an energy-efficient approach for task scheduling in vehicular edge computing networks based on DRL,with the purpose of both guaranteeing the task latency requirement for multiple users and minimizing total energy consumption of MEC servers.Numerical results demonstrate that the proposed algorithm outperforms other methods.
文摘The zone of proximal development(ZPD) and the scaffolding theory are very different,both in terms of their theoretical origins and connotations,and can even be said to be very different.However,during the development of the two concepts,some scholars have misunderstood them,resulting in the two being mistaken for similar concepts and therefore often confused.Professor James Lantolf from Pennsylvania State University(State College,USA) was interviewed by Professor Lili Qin from Dalian University of Foreign Studies(Dalian,China) and provides an indepth analysis of these issues.The interview begins with the theoretical roots,connotations and definitions of the ZPD and scaffolding concepts,and then unravels the story of how they have been“mistakenly loved for life”,and ultimately it is made clear that the two concepts are completely different in their practical application to language teaching and should not continue to be used interchangeably.
基金National Natural Science Foundation of China,Grant/Award Numbers:61703418,61825305。
文摘Here,the challenges of sample efficiency and navigation performance in deep rein-forcement learning for visual navigation are focused and a deep imitation reinforcement learning approach is proposed.Our contributions are mainly three folds:first,a frame-work combining imitation learning with deep reinforcement learning is presented,which enables a robot to learn a stable navigation policy faster in the target-driven navigation task.Second,the surrounding images is taken as the observation instead of sequential images,which can improve the navigation performance for more information.Moreover,a simple yet efficient template matching method is adopted to determine the stop action,making the system more practical.Simulation experiments in the AI-THOR environment show that the proposed approach outperforms previous end-to-end deep reinforcement learning approaches,which demonstrate the effectiveness and efficiency of our approach.
文摘Providing autonomous systems with an effective quantity and quality of information from a desired task is challenging. In particular, autonomous vehicles, must have a reliable vision of their workspace to robustly accomplish driving functions. Speaking of machine vision, deep learning techniques, and specifically convolutional neural networks, have been proven to be the state of the art technology in the field. As these networks typically involve millions of parameters and elements, designing an optimal architecture for deep learning structures is a difficult task which is globally under investigation by researchers. This study experimentally evaluates the impact of three major architectural properties of convolutional networks, including the number of layers, filters, and filter size on their performance. In this study, several models with different properties are developed,equally trained, and then applied to an autonomous car in a realistic simulation environment. A new ensemble approach is also proposed to calculate and update weights for the models regarding their mean squared error values. Based on design properties,performance results are reported and compared for further investigations. Surprisingly, the number of filters itself does not largely affect the performance efficiency. As a result, proper allocation of filters with different kernel sizes through the layers introduces a considerable improvement in the performance.Achievements of this study will provide the researchers with a clear clue and direction in designing optimal network architectures for deep learning purposes.
文摘This paper reports on a study on the effects of reading-writing integrated tasks on vocabulary learning and explored the differential roles of creative construction and non-creative construction in promoting lexical learning. Participants were 90 first-year English majors, randomly assigned to two experimental groups(continuation and retelling) and one control group, with 30 students in each group. Results showed that the continuation group generated a substantial amount of creative construction and produced significantly more instances of creative imitation than the retelling group. The continuation group outperformed the retelling group for both receptive and productive vocabulary knowledge gain and retention, but differences were only significant in terms of productive vocabulary retention. Finally, productive vocabulary knowledge retention among the continuation group was significantly and positively correlated with creative imitation(meaning creation coupled with language imitation), but not with linguistic alignment per se. As productive vocabulary knowledge constitutes the learner ’s ability to use lexical knowledge to express ideas in dynamic contexts, the findings afforded evidence that creative imitation could be the answer to the fundamental issue of L2 learning(i.e., mapping static language onto dynamic idea expression). The pedagogical implications as well as future research directions are also discussed.
文摘Night,Mother is a famous play about female identification.It is about Jessie,a middle aged woman,who,through many years’struggle,decides to kill herself at nine o’clock in an ordinary night in her mother’s house.Jessie’s fighting for female identity and corresponding power would be analyzed.In the play,Jessie tries two ways to get her power,one by trying to imitating men to get accepted by patriarchal society,and the other by revolting outside resistance and insisting her pursuit of self rights and freedom.
