139 results about "Decision graph" patented technology

Methods and systems are disclosed for learning a regression decision graph model using a Bayesian model selection approach. In a disclosed aspect, the model structure and/or model parameters can be learned using a greedy search algorithm applied to grow the model so long as the model improves. This approach enables construction of a decision graph having a model structure that includes a plurality of leaves, at least one of which includes a non-trivial linear regression. The resulting model thus can be employed for forecasting, such as for time series data, which can include single or multi-step forecasting.

The invention relates to a road network trajectory clustering-based travel hotspot region mining method. According to the method, taxi trajectories are mapped into a road network, and a clustering method combining points of interest and trajectories collected in actual roads is adopted; on the basis of a density peak clustering algorithm, an OPAM algorithm which optimizes an initial center based on density peaks is provided, namely, a namely DP-OPAM algorithm is provided; according to the algorithm, the local density of data points and shortest distances from the points to points with higher density are adopted, and a decision graph is adopted to select the category of data points with higher density and closest distance, and the category is adopted as an initial clustering center; and onthe basis of the initial clustering center, an OPAM clustering algorithm additionally adopting inverse learning is used to obtain a clustering result. Compared with an original OPAM algorithm, the newalgorithm can not only automatically determine a clustering center, but also improve accuracy and shorten clustering time and realize the analysis of user travel hotspots.

The present invention utilizes a cross-prediction scheme to predict values of discrete and continuous time observation data, wherein conditional variance of each continuous time tube variable is fixed to a small positive value. By allowing cross-predictions in an ARMA based model, values of continuous and discrete observations in a time series are accurately predicted. The present invention accomplishes this by extending an ARMA model such that a first time series “tube” is utilized to facilitate or “cross-predict” values in a second time series tube to form an “ARMAxp” model. In general, in the ARMAxp model, the distribution of each continuous variable is a decision graph having splits only on discrete variables and having linear regressions with continuous regressors at all leaves, and the distribution of each discrete variable is a decision graph having splits only on discrete variables and having additional distributions at all leaves.

The present invention provides collaborative filtering systems and methods employing default scores of decision graphs/trees to quickly create a top-n prediction list that can efficiently determine a user's interest in items. In one instance of the present invention, the list is refined by utilizing a variable maximum score algorithm and/or an invalidation list algorithm to insert items that score above an inclusion threshold set by a last item in the top-n prediction list. In another instance of the present invention, an invalidation list for a decision graph and/or decision tree is utilized to create a top-n prediction list. An algorithm employing default scores is then utilized to refine the top-n prediction list to insert items with default scores above an inclusion threshold set by a last item in the top-n prediction list.

Embodiments of the invention disclose a density peak-based big data mining method and apparatus. The method comprises a first stage of selecting initial clustering centers by using a density peak clustering algorithm, and a second stage of determining an initial clustering number and accelerating convergence of a conventional fuzzy C-mean algorithm. In the algorithm of the first stage, a decision graph is provided by using density peak clustering, the initial clustering centers are selected, and the initial clustering center number is namely a clustering center number after the selection is finished, so that the clustering number does not need to be manually specified. In the algorithm of the second stage, an iterative process of the conventional fuzzy C-mean algorithm is optimized, the influence of local density on the algorithm is considered, and a density weighting factor is added, so that the algorithm can obtain a global optimal solution in an accelerated way; and an oscillating factor is added, so that the algorithm convergence can be accelerated.

A method, apparatus and computer-readable medium for conjunctive binary decision diagram building and variable quantification using case-splitting are presented. A BDD building program builds a BDD for at least one node in a netlist graph representation of a circuit design. One or more variables are selected for case-splitting. The variable is set to a constant logical value and then the other. A BDD is built for each case. The program determines whether the variable is scheduled to be quantified out. If so, the program combines the BDDs for each case according to whether the quantification is existential or universal. If the variable is not scheduled to be quantified, the program combines the BDDs for each case so that the variable is introduced back into the resulting BDD, which has a reduced number of peak live nodes.

The invention relates to a fast multi-focus image fusion algorithm. The algorithm comprises the following steps of firstly, decomposing a source image by using a smoothing filter and constructing thedecision graph model of multi-focus image fusion; secondly, using the correlation between the neighborhood pixels of multi-focus images to provide a fast steering filter optimization decision graph sothat calculating complexity is greatly reduced; then, reconstructing two scales of images; and finally, using the fusion method of the invention and other triple existing fusion methods to fuse the multi-focus images. Compared with the other three kinds of fusion methods, by using the method of the invention, regardless of subjective visual and objective evaluation indexes, the texture information, the significance information and other details of the multi-focus images can be preserved.

The invention discloses a contact network failure risk assessment method based on a binary decision graph algorithm, comprising the steps of: (1) generating a BDD structure, determining system boundaries, basic events and top events, establishing and normalizing a fault tree, and generating a BDD structure, wherein the corresponding BDD nodes may be directly created by ITE operations for basic events; and ITE operations may be performed on basic events or other intermediate events to obtain the BDD structure of the original intermediate event for intermediate events; (2) calculating the accident rate of the contact network failure risk, generating a Boolean logic expression by the fault tree, and generating a Boolean logic function corresponding to the fault tree top event, wherein when the true value is obtained, the probability of occurrence of the top event or any intermediate event may be obtained; and (3) measuring the event importance. The invention applies the BDD method to thefailure risk assessment of the contact network, which simplifies the calculation process, and solves the problems such as the combined explosion and the complicated solving process encountered by thecut set method in the contact network failure fault tree analysis.

The invention discloses an automatic identification method of microaneurysm in a fundus colorful image. The method comprises the following steps: firstly, carrying out preprocessing and shade correction on an image of a green channel, then, utilizing the local information of a sliding window to determine an adaptive threshold value to carry out binarization on the image, obtaining a decision graph through a double-window filter based on connectivity restriction, and carrying out region growing on the decision graph, wherein the result of the region growing is the accurate identification microaneurysm. By use of the whole method, the fine identification and segmentation of the microaneurysm can be realized, identification missing and misjudgement can be effectively avoided, the processes of feature extraction and classifier classification do not need to be adopted, and identification efficiency is effectively improved.

The invention provides a feedback density peak value clustering method and a system thereof. A problem that one type is divided into a plurality of types when a plurality of density peak values are generated in one type in an original density peak value algorithm is solved. Simultaneously, accuracy of an original algorithm in a high dimension data set is increased. The method comprises the following steps of 1, using non-negative matrix factorization to carry out characteristic extraction on a data set; 2, according to an original density peak value clustering algorithm, drawing a decision graph and selecting a plurality of clustering centers; 3, using a ''nearest neighbor'' algorithm to distribute the rest of points and removing a noise point; 4, using a SVM to feed back a clustering result between two types; and 5, according to a feedback result, merging the types which can be merged. By using the method, robustness of a density peak value algorithm can be effectively increased, clusters with any shapes can be well discovered, high dimension data can be effectively processed, and a good clustering effect is possessed.

The present invention provides a method and mechanism for simulating complex digital circuits using hybrid control and data flow representations. Specifically, the invention provides a method of simulating a digital circuit in such a way that the simulation is stopped at desired functions for subsequent analysis. A hardware design code describing the digital circuit is converted to an assignment decision diagram (ADD) representation that is then annotated with one or more control nodes that are used for maintaining control flow through a simulator. In this way, one or more break points are created that allow the simulator to stop at associated points in the simulation.

The embodiment of the invention discloses a text clustering method, device and equipment based on density and a storage medium, and relates to the technical field of text data analysis. The method comprises the steps of receiving a target data set; determining a target distance formula; generating a distance matrix about the whole target data set; calculating the local density of each data point;separately extracting the minimum value of the distance between each data point and each data point in the sample point set, and recording the minimum value as the minimum point distance; establishinga clustering decision diagram according to the local density and the minimum point distance; determining the number of class clusters and a class cluster center in the clustering decision diagram; and dividing each data point into class clusters of the clustering decision diagram. According to the method, in the whole clustering process, the non-spherical data can be clustered only by calculatingthe distance between the sample points once without iterative calculation, the algorithm performance is greatly improved, the clustering decision diagram is used for scientifically selecting the number of the class clusters, and the situation that the number of the class clusters is manually set without basis is avoided.

A method for verifying a design through symbolic simulation is disclosed. The method comprises creating one or more binary decision diagram variables for one or more inputs in a design containing one or more state variables and building a binary decision diagram for a first node of one or more nodes of the design. A binary decision diagram for the initial state function of one or more state variables of the design is generated and the design is subsequently initialized. Binary decisions diagrams for one or more constraints are synthesized. A set of constraint values is accumulated over time by combining the binary decision diagrams for the one or more constraints with a set of previously generated binary decision diagrams for a set of constraints previously used in one or more previous time-steps. A binary decision diagram for the next state function of the one or more state variables in the design is constructed in the presence of the constraints. The one or more state variables in the design are updated by propagating the binary decision diagram for the next state function to the one or more state variables and a set of binary decision diagrams for the one or more targets in the presence of the one or more constraints is calculated. The set of binary decision diagrams for one or more targets is constrained and the design is verified by determining whether the one or more targets were hit.

The invention provides a network test method and system, a storage medium and electronic equipment, and the method comprises the steps: obtaining a target binary decision diagram matched with a flow table rule and a P4 program; generating and updating a probe according to the target binary decision diagram, the topology and the rule updating event; forwarding the probe, and writing an action corresponding to the flow table rule into the probe; and when the probe returns to the server, checking the probe to obtain a detection result of a rule error in the operation process. According to the network test method and system, the storage medium and the electronic equipment provided by the embodiment of the invention, the binary decision diagram can be used for representing the P4 program and the flow table rule, so that the low overhead of network resources is realized, and the full coverage of the flow table rule can be realized.

The invention discloses a clustering method and device based on a density peak value, and the method comprises the steps: randomly selecting a plurality of observation points in a sample space; obtaining a clustering result of each observation point according to the distance between each sample point and each observation point; generating a similarity matrix of high-dimensional data according to the clustering results of all observation points; enabling the sum of all rows of the similarity matrix to serve as the density of sample points corresponding to each row of the similarity matrix; obtaining a density set of the similarity matrix according to the density of sample points corresponding to each row of the similarity matrix; calculating the peak value of each density in the density set; determining a candidate center according to the peak value of each density; carrying out the clustering of the peak values of the candidate centers, and obtaining the clustering result of each sample point. According to the invention, there is no need of the setting of cutoff distance parameters and there is no need of a decision making map for the manual selection of clustering centers, so themethod is wider in application range.