In many machine learning applications,data are not free,and there is a test cost for each data item. For the economical reason,some existing works try to minimize the test cost and at the same time,preserve a particul...In many machine learning applications,data are not free,and there is a test cost for each data item. For the economical reason,some existing works try to minimize the test cost and at the same time,preserve a particular property of a given decision system. In this paper,we point out that the test cost one can afford is limited in some applications. Hence,one has to sacrifice respective properties to keep the test cost under a budget. To formalize this issue,we define the test cost constraint attribute reduction problem,where the optimization objective is to minimize the conditional information entropy. This problem is an essential generalization of both the test-cost-sensitive attribute reduction problem and the 0-1 knapsack problem,therefore it is more challenging. We propose a heuristic algorithm based on the information gain and test costs to deal with the new problem. The algorithm is tested on four UCI（University of California-Irvine） datasets with various test cost settings. Experimental results indicate the appropriate setting of the only user-specified parameter λ.展开更多
This paper presents an approach for extending the constraint model defined for conformity testing of a given method of class to its overriding method in subclass using inheritance principle. The first objective of the...This paper presents an approach for extending the constraint model defined for conformity testing of a given method of class to its overriding method in subclass using inheritance principle. The first objective of the proposed work is to find the relationship between the test model of an overriding method and its overridden method using the constraint propagation. In this context the approach shows that the test cases developed for testing an original method can be used for testing its overriding method in a subclass and then the number of test cases can be reduced considerably. The second objective is the use of invalid data which do not satisfy the precondition constraint and induce valid output values for introducing a new concept of test called secure testing. The implementation of this approach is based on a random generation of test data and analysis by formal proof.展开更多
Advancements in semiconductor technology are making gate-level test generation more challenging. This is because a large amount of detailed structural information must be processed in the search process of automatic t...Advancements in semiconductor technology are making gate-level test generation more challenging. This is because a large amount of detailed structural information must be processed in the search process of automatic test pattern generation (ATPG). In addition, ATPG needs to deal with new defects caused by process variation when IC is shrinking. To reduce the computation effort of ATPG, test generation could be started earlier at higher abstraction level, which is in line with top-down design methodology that has become more popular nowadays. In this research, we employ Chen’s high-level fault model in the high-level ATPG. Besides shorter ATPG time as shown in many previous works, our study showed that high-level ATPG also contributes to test compaction. This is because most of the high-level faults correlate with the gate-level collapsed faults especially at input/output of the modules in a circuit. The high-level ATPG prototype used in our work is mainly composed by constraint-driven test generation engine and fault simulation engine. Experimental result showed that more reduced/compact test set can be generated from the high-level ATPG.展开更多
Image segmentation is a key and fundamental problem in image processing,computer graphics,and computer vision.Level set based method for image segmentation is used widely for its topology flexibility and proper mathem...Image segmentation is a key and fundamental problem in image processing,computer graphics,and computer vision.Level set based method for image segmentation is used widely for its topology flexibility and proper mathematical formulation.However,poor performance of existing level set models on noisy images and weak boundary limit its application in image segmentation.In this paper,we present a region consistency constraint term to measure the regional consistency on both sides of the boundary,this term defines the boundary of the image within a range,and hence increases the stability of the level set model.The term can make existing level set models significantly improve the efficiency of the algorithms on segmenting images with noise and weak boundary.Furthermore,this constraint term can make edge-based level set model overcome the defect of sensitivity to the initial contour.The experimental results show that our algorithm is efficient for image segmentation and outperform the existing state-of-art methods regarding images with noise and weak boundary.展开更多
R-DSP(Radar Digital Signal Processor)芯片中BSU(Branch Shift Unit)运算部件具有较大的设计规模和复杂度,传统Verilog验证平台难以满足其验证需求问题。针对该问题,文中采用UVM(Universal Verification Methodology)方法对BSU运算部...R-DSP(Radar Digital Signal Processor)芯片中BSU(Branch Shift Unit)运算部件具有较大的设计规模和复杂度,传统Verilog验证平台难以满足其验证需求问题。针对该问题,文中采用UVM(Universal Verification Methodology)方法对BSU运算部件进行功能验证。搭建基于SystemVerilog语言实现的UVM验证平台,使用定向测试和带约束的随机测试进行验证,并采用覆盖率驱动的方法指导测试用例的生成,以充分覆盖BSU运算部件的各个功能和代码路径。经过多轮测试激励验证,代码覆盖率接近100%,完成了对BSU运算部件的功能验证。所提方法为R-DSP芯片中的ALU(Arithmetic Logic Unit)、AGU(Address Generation Unit)、MU(Multiplication Unit)等运算部件的验证工作提供了参考和借鉴。展开更多
Optimum utilization of the loading capability of engineering materials is an important and active contribution to protect nature's limited resources,and it is the key for economic design methods.In order to make u...Optimum utilization of the loading capability of engineering materials is an important and active contribution to protect nature's limited resources,and it is the key for economic design methods.In order to make use of the materials' resources,those must be known very well;but conventional test methods will offer only limited informational value.The range of questions raised is as wide as the application of engineering materials,and partially they are very specific.The development of huge computer powers enables numeric modelling to simulate structural behaviour in rather complex loading environments-so the real material behaviour is known under the given loading conditions.Here the art of material testing design starts.To study the material behaviour under very distinct and specific loading conditions makes it necessary to simulate different temperature ranges,loading speeds, environments etc.and mostly there doesn't exist any commonly agreed test standard.In this contribution two popular,non-standard test procedures and test systems will be discussed on the base of their application background,special design features as well as test results and typically gained information:The demand for highspeed tests up to 1000 s^(-1) of strain rate is very specific and originates primarily in the automotive industry and the answers enable CAE analysis of crashworthiness of vehicle structures under crash conditions.The information on the material behaviour under multiaxial loading conditions is a more general one.Multiaxial stress states can be reduced to an equivalent stress,which allows the evaluation of the material's constraint and criticality of stress state.Both discussed examples shall show that the open dialogue between the user and the producer of testing machines allows custom-tailored test solutions.展开更多
基金supported by the National Natural Science Foundation of China under Grant No. 60873077/F020107
文摘In many machine learning applications,data are not free,and there is a test cost for each data item. For the economical reason,some existing works try to minimize the test cost and at the same time,preserve a particular property of a given decision system. In this paper,we point out that the test cost one can afford is limited in some applications. Hence,one has to sacrifice respective properties to keep the test cost under a budget. To formalize this issue,we define the test cost constraint attribute reduction problem,where the optimization objective is to minimize the conditional information entropy. This problem is an essential generalization of both the test-cost-sensitive attribute reduction problem and the 0-1 knapsack problem,therefore it is more challenging. We propose a heuristic algorithm based on the information gain and test costs to deal with the new problem. The algorithm is tested on four UCI（University of California-Irvine） datasets with various test cost settings. Experimental results indicate the appropriate setting of the only user-specified parameter λ.
文摘This paper presents an approach for extending the constraint model defined for conformity testing of a given method of class to its overriding method in subclass using inheritance principle. The first objective of the proposed work is to find the relationship between the test model of an overriding method and its overridden method using the constraint propagation. In this context the approach shows that the test cases developed for testing an original method can be used for testing its overriding method in a subclass and then the number of test cases can be reduced considerably. The second objective is the use of invalid data which do not satisfy the precondition constraint and induce valid output values for introducing a new concept of test called secure testing. The implementation of this approach is based on a random generation of test data and analysis by formal proof.
文摘Advancements in semiconductor technology are making gate-level test generation more challenging. This is because a large amount of detailed structural information must be processed in the search process of automatic test pattern generation (ATPG). In addition, ATPG needs to deal with new defects caused by process variation when IC is shrinking. To reduce the computation effort of ATPG, test generation could be started earlier at higher abstraction level, which is in line with top-down design methodology that has become more popular nowadays. In this research, we employ Chen’s high-level fault model in the high-level ATPG. Besides shorter ATPG time as shown in many previous works, our study showed that high-level ATPG also contributes to test compaction. This is because most of the high-level faults correlate with the gate-level collapsed faults especially at input/output of the modules in a circuit. The high-level ATPG prototype used in our work is mainly composed by constraint-driven test generation engine and fault simulation engine. Experimental result showed that more reduced/compact test set can be generated from the high-level ATPG.
基金supported in part by the NSFC-Zhejiang Joint Fund of the Integration of Informatization and Industrialization(U1609218)NSFC(61772312,61373078,61772253)+1 种基金the Key Research and Development Project of Shandong Province(2017GGX10110)NSF of Shandong Province(ZR2016FM21,ZR2016FM13)
文摘Image segmentation is a key and fundamental problem in image processing,computer graphics,and computer vision.Level set based method for image segmentation is used widely for its topology flexibility and proper mathematical formulation.However,poor performance of existing level set models on noisy images and weak boundary limit its application in image segmentation.In this paper,we present a region consistency constraint term to measure the regional consistency on both sides of the boundary,this term defines the boundary of the image within a range,and hence increases the stability of the level set model.The term can make existing level set models significantly improve the efficiency of the algorithms on segmenting images with noise and weak boundary.Furthermore,this constraint term can make edge-based level set model overcome the defect of sensitivity to the initial contour.The experimental results show that our algorithm is efficient for image segmentation and outperform the existing state-of-art methods regarding images with noise and weak boundary.
文摘Optimum utilization of the loading capability of engineering materials is an important and active contribution to protect nature's limited resources,and it is the key for economic design methods.In order to make use of the materials' resources,those must be known very well;but conventional test methods will offer only limited informational value.The range of questions raised is as wide as the application of engineering materials,and partially they are very specific.The development of huge computer powers enables numeric modelling to simulate structural behaviour in rather complex loading environments-so the real material behaviour is known under the given loading conditions.Here the art of material testing design starts.To study the material behaviour under very distinct and specific loading conditions makes it necessary to simulate different temperature ranges,loading speeds, environments etc.and mostly there doesn't exist any commonly agreed test standard.In this contribution two popular,non-standard test procedures and test systems will be discussed on the base of their application background,special design features as well as test results and typically gained information:The demand for highspeed tests up to 1000 s^(-1) of strain rate is very specific and originates primarily in the automotive industry and the answers enable CAE analysis of crashworthiness of vehicle structures under crash conditions.The information on the material behaviour under multiaxial loading conditions is a more general one.Multiaxial stress states can be reduced to an equivalent stress,which allows the evaluation of the material's constraint and criticality of stress state.Both discussed examples shall show that the open dialogue between the user and the producer of testing machines allows custom-tailored test solutions.
