The utilization of expert knowledge to construct Bayesian Network (BN) is one of the key issues of uncertain inference. In this paper the learning procedure of BN is simplified by applying data dependencies in relatio...
详细信息
The utilization of expert knowledge to construct Bayesian Network (BN) is one of the key issues of uncertain inference. In this paper the learning procedure of BN is simplified by applying data dependencies in relational database, which help to remove extraneous attributes and build the initial network structure. The final classification model named FM-NB inherits the simplicity of Naive Bayes (NB) and the ability to maintain inter-dependencies among attributes of Tree augmented NB(TAN). The inter-dependencies are expressed by Functional dependencies (FDs) and Local Multivalued dependencies (LMVDs), which relaxed the precondition about attribute sets of Multivalued dependency (MVD) and solved the problem related to conditional independence of Embedded Multivalued dependencies (EMVDs). Experimental results are presented to show the effectiveness and efficiency of the proposed approach.
Given a multi-features data set, a best preference query (BPQ) computes the maximal preference score (MPS) that the tuples in the data set can achieve with respect to a preference function. BPQs are very useful in app...
详细信息
Convex evidence theory model is the only method to fuse ordered proposition, in this paper we proposed a novel convex evidence theory model based on Gaussian function, we modified Gaussian function and used it to comb...
详细信息
Convex evidence theory model is the only method to fuse ordered proposition, in this paper we proposed a novel convex evidence theory model based on Gaussian function, we modified Gaussian function and used it to combine mass function of ordered propositions, we presented the calculation of the parameters of this function, and proposed a more accurate method to find the most likely true proposition. Theoretical analysis and experimental results demonstrate that the proposed method has lower time complexity and higher accuracy in some cases than existing methods.
In this paper, a new security requirement about RFID protocols is proposed firstly, it is lightweight encryption function structure protection. Subsequently, a lightweight RFID authentication protocol is proposed as a...
详细信息
In this paper, a new security requirement about RFID protocols is proposed firstly, it is lightweight encryption function structure protection. Subsequently, a lightweight RFID authentication protocol is proposed as a case study, this protocol only requires O(1) work to identify and authenticate a tag in the server and only one-way hash function, XOR operation are needed in tags. In addition, we give formal reasoning for this protocol by using GNY logic;the security properties of this protocol are analyzed as well by comparing with the related RFID protocols.
In formal verification, the first step is to build a formal model. Most of the academic tools represent the designs in BLIF or BLIF-MV network, while cannot directly take Verilog source code as an input. For instance,...
详细信息
In formal verification, the first step is to build a formal model. Most of the academic tools represent the designs in BLIF or BLIF-MV network, while cannot directly take Verilog source code as an input. For instance, VIS depends on the translator vl2mv to compile Verilog into the intermediate format BLIF-MV first, and then extracts the FSM (Finite State Models) from it instead of the direct Verilog source code. However, Verilog and VHDL are the most common used hardware description languages for large scale integrated circuits. It is necessary if the source code could be used directly, rather than an intermediate format. We notice that synthesizable Verilog is well-structured which can be represented in SSA (Static Single Assignment) form. In this paper, we describe a method that extracts formal model directly from Verilog source code by converting the code to SSA form. With this method, we have implemented a prototype system that extracts formal model of Verilog programs. We also prove the conversion function-preserving by translating SSA back to Verilog and checking its equivalence with the primary design. To ensure correctness and robustness, we have tested our system with ISCAS and picoJava-II, a specification of microprocessor, and the experimental results on these realistic circuits show that the conversion procedure is function-preserving. Moreover, we take it as the front-end of CEC (Combinational Equivalence Checking) and demonstrate the formal model extracted with our method can be exploited efficiently in CEC.
Except common security and privacy requirements, well scalability and supporting tag ownership transfer are important requests for RFID systems. In this paper, an efficient RFID mutual authentication protocol supporti...
详细信息
Except common security and privacy requirements, well scalability and supporting tag ownership transfer are important requests for RFID systems. In this paper, an efficient RFID mutual authentication protocol supporting tag ownership transfer is proposed, this protocol only requires O(1) work to identify and authenticate a tag in the backend server and is suitable for the low-cost RFID systems. The security and performance of the proposed protocol are analyzed as well.
The purpose of this paper is to study the topological relations of four simple regions. We extend 4-intersection matrix model to 16-intersections matrix model to represent the spatial relations of four simple regions,...
详细信息
The purpose of this paper is to study the topological relations of four simple regions. We extend 4-intersection matrix model to 16-intersections matrix model to represent the spatial relations of four simple regions, and prove that these topological relations are exclusive and complete. Moreover, we obtained 32406 topological relations in practice by programs, and we illustrate that the 16-intersections matrix model has stronger expressive ability than the topological relations given by RCC-5 in pairs.
This paper presents an efficient technique for processing of 3D meshed surfaces via spherical wavelets. More specifically, an input 3D mesh is firstly transformed into a spherical vector signal by a fast low distortio...
详细信息
This paper presents an efficient technique for processing of 3D meshed surfaces via spherical wavelets. More specifically, an input 3D mesh is firstly transformed into a spherical vector signal by a fast low distortion spherical parameterization approach based on symmetry analysis of 3D meshes. This signal is then sampled on the sphere with the help of an adaptive sampling scheme. Finally, the sampled signal is transformed into the wavelet domain according to spherical wavelet transform where many 3D mesh processing operations can be implemented such as smoothing, enhancement, compression, and so on. Our main contribution lies in incorporating a fast low distortion spherical parameterization approach and an adaptive sampling scheme into the frame for pro- cessing 3D meshed surfaces by spherical wavelets, which can handle surfaces with complex shapes. A number of experimental ex- amples demonstrate that our algorithm is robust and efficient.
In this paper, based on the newly introduced mappings ξi (x, yi) ∈ Rn, i = 1, , m and ηi (x, zj) ∈ Rn, j = 1, , l, using the idea of homotopy methods, we propose a homotopy interior point method for solving the op...
详细信息
An imbalanced data classification algorithm named as SBoost is proposed. SBoost regards the decision tree as a weak classifier. Only the instances that their prediction value is more than the threshold are updated wit...
详细信息
An imbalanced data classification algorithm named as SBoost is proposed. SBoost regards the decision tree as a weak classifier. Only the instances that their prediction value is more than the threshold are updated with their weights in every iteration. At the beginning of each iteration, SBoost makes use of data generation method to add synthetic minority class instances in order to balance training information. After the sub-classifier is formed, the algorithm deletes the synthetic instances that are not correctly classified. This paper puts forwards the theoretical analysis of training error bound. The experimental results show that SBoost has advantages on imbalanced data classification problem.
暂无评论