112 results about "Cost sensitive" patented technology

An automated energy rate reduction and demand side sequencing management and analysis system bridges the gap between supply and demand side energy management. The management system enables energy consumer's to determine, automate and react in “real-time” to all of the cost sensitive energy billing components in a unregulated or regulated utility energy supplier rate as well as determine a “real-time” demand side operational sequence in order to drive new costs in their facility.

Record Linkage (RL) is the task of identifying two or more records referring to the same entity (e.g., a person, a company, etc.). RL models can be based on Cost Sensitive Alternating Decision Trees (ADTree), an algorithm that uniquely combines boosting and decision trees algorithms to create shorter and easier-to-interpret linking rules. These models can be naturally trained to operate at industrial precision/recall operating points, and the shorter output rules are so clear that it can effectively explain its decisions to non-technical users via score aggregation or visualization. The models significantly outperform other baselines on the desired industrial operating points, and the improved understanding of the model's decisions led to faster debugging and feature development cycles.

A means for extending a PCI System of a host computer via software-centric virtualization. A Root Complex is virtualized at the host computer, and physically separated with a portion located remotely at an Endpoint, such as at a Remote Bus Adapter. One aspect of the invention avoids the need for a Host Bus Adapter. The invention utilizes 1 Gbps-10 Gbps or greater connectivity via the host's existing standard LAN adapter along with unique software to form the virtualization solution. The invention works within a host's PCI Express topology, extending the topology by adding an entire virtual I/O hierarchy via virtualization. The invention enables I/O virtualization in those implementations where a specialized host bus may not be desirable or feasible. Some examples of this may be a laptop computer, an embedded design, a cost-sensitive design, or a blade host where expansion slots are not available or accessible.

A network security apparatus includes a packet detector detecting transmission of data packets between a plurality of hosts and a plurality of domains and defining a plurality of links therefrom. A model builder circuit receives the plurality of links from the packet detector, receives ground truth information labeling one or more of the plurality of hosts or one or more of the plurality of domains as benign or malicious, generates predictive models from the received links and ground truth information, and stores generated predictive models in a predictive model database. An anomaly detector circuit retrieves the generated predictive models from the predictive model database and uses the predictive models to label each of the plurality of hosts and plurality of domains, that have not previously been labeled by the ground truth information, as benign or malicious.

A pipelined data processing unit includes an instruction sequencer and n functional units capable of executing n operations in parallel. The instruction sequencer includes a random access memory for storing very-long-instruction-words (VLIWs) used in operations involving the execution of two or more functional units in parallel. Each VLIW comprises a plurality of short-instruction-words (SIWs) where each SIW corresponds to a unique type of instruction associated with a unique functional unit. VLIWs are composed in the VLIW memory by loading and concatenating SIWs in each address, or entry. VLIWs are executed via the execute-VLIW (XV) instruction. The iVLIWs can be compressed at a VLIW memory address by use of a mask field contained within the XV1 instruction which specifies which functional units are enabled, or disabled, during the execution of the VLIW. The mask can be changed each time the XV1 instruction is executed, effectively modifying the VLIW every time it is executed. The VLIW memory (VIM) can be further partitioned into separate memories each associated with a function decode-and-execute unit. With a second execute VLIW instruction XV2, each functional unit's VIM can be independently addressed thereby removing duplicate SIWs within the functional unit's VIM. This provides a further optimization of the VLIW storage thereby allowing the use of smaller VLIW memories in cost sensitive applications.

The invention discloses an unbalanced data integration classification method that combines resampling technology and cost-sensitive learning, relates to the field of artificial intelligence integrated learning, and mainly solves the problem of unbalanced data classification using complete data information in the prior art. The method The steps are: (1) input the training data set; (2) calculate the relative density of the sample space distribution; (3) resample to generate multiple subsets and train the basic classifier; (4) calculate the similarity matrix of the test sample; (5) ) using multi-objective optimization and integration to obtain prior results; (6) performing cost-sensitive learning prediction on the test set; (7) using KL divergence to optimize and fuse the results. The method designs a new sampling method to solve the problem of unbalanced data distribution; uses a method combining resampling technology and cost-sensitive learning to solve the problem of incomplete information; and makes full use of the data information of the test set itself to improve integration performance of the classifier.

A technique is provided for acquiring desired information from one or more information sources. The technique includes assessing the one or more information sources for the desired information, searching for the desired information within the one or more information sources based on the assessment, and automatically acquiring the desired information based on the search.

The invention provides a cost-sensitive stacking integrated learning framework based on feature inverse mapping in order to effectively solve the problem of unbalanced classification. Firstly, a random forest, a limit forest, a gradient tree, linear discriminant analysis and logistic regression are simultaneously adopted for training of data sets as basic classifiers; then confidences obtained bycross validation of the basic classifiers are stacked through a stacking integrated learning method to form a new feature set; feature exponential transform of the new feature set is performed, an exponent shown in the description of the optimal average logarithmic loss is selected, and feature inverse mapping of the feature is performed with the exponent shown in the description; and finally, thefeature set after inverse mapping is trained by employing cost-sensitive logistic regression. In test steps, the feature obtained by stacking avoids the operation of inverse mapping. Compared with aconventional unbalanced classification integration method, according to the cost-sensitive stacking integrated learning framework, cost sensitivity and stacking integration are firstly combined so that the generalization performance of the unbalanced classification problem is effectively enhanced, and a model can obtain a stable classification threshold.

The invention provides a breast tumor diagnosis system based on magnetic resonance spectrum imaging, comprising a superconduction MR (Magneto Resistance) scanner and a computer system. Through a computer auxiliary detection measure, the breast tumor diagnosis system realizes the early identification diagnosis of a breast tumor and the further diagnosis of breast cancer. Mammary gland magnetic resonance spectrum data are studied by adopting a distinctive manifold study method and projected to a low dimension embedding space, thus not only a low dimension manifold structure hidden in a high dimension magnetic resonance spectrum space can be revealed, but also the identification information in the mammary gland magnetic resonance spectrum data can be efficiently kept; then optimization clustering is carried out on the low dimension identification characteristic by utilizing a clustering method so that data points without the similar identification characteristic are separated to the maximum; and further a cost sensitive mechanism is introduced to achieve misclassification total cost minimization and realize optimization diagnosis of the breast cancer.

The invention discloses a boosting based cost sensitive software defect prediction method, belonging to the technical field of software engineering application. According to the invention, re-sampling is carried out by means of Bootstrap, the erroneous deletion of valuable attributes can be prevented by using a cost-sensitive subclass selection manner of randomly deleting an attribute during attribute selection, and meanwhile, the selected attribute sub-class facilitates the reduction of prediction error costs, when the weight is updated, a cost-sensitive weight updating mechanism is adopted, large weight is endowed to a data set with high cost, so that the data can be guaranteed to be studied for many times to obtain a reasonable integrated prediction model, and the prediction model is applied to small sample data to accurately predict software defects, and therefore, the technical problems that the prediction effects are not satisfactory due to the shortage of training data under the small sample data and the unequal cost of false positives and false negatives in the prediction process are solved.

A flowmeter is provided with a solid-state sensor technology for increasing the accuracy and efficiency of fluid flow monitoring and leak detection in the irrigation industry and other cost-sensitive flow-critical applications. The device can monitor and detect usage of fluids such as water more accurately and economically than presently possible to help eliminate waste and ensure correct billing. Measurement can be over an optimized narrow range or can be automatically switched to cover consecutive ranges from leakage to gross fluid conduit faults. With its minimal component count, ease of assembly, independence from the specifics of its housing, low maintenance needs, and its interchangeability this device can serve a wide range of metering needs from low end leak monitoring to standard flow measurement.

The invention belongs to a computational biology information technique, and particularly relates to a method for predicting epitope through cost-sensitive integrating and clustering on the basis of a sequence. The method comprises the main steps that 1, descriptive features of antigen protein residues are constructed, wherein the features comprise the evolutionary conservation feature, the secondary structure feature, the disordered region feature, the dipeptide composition feature and physical and chemical attributes; 2, an optimal feature subset is selected through Fisher-Markov and an incremental iterative feature selection method; 3, unbalanced data sets are processed through cost-sensitive integrating learning; 4, potential epitope residues are predicted from antigenic determination residues through a spatial clustering algorithm. The method is suitable for antigen protein epitope prediction of known and unknown structure information and is also suitable for large-scale application and popularization.

A method (and structure) for processing an inductive learning model for a dataset of examples, includes dividing the dataset into N subsets of data and developing an estimated learning model for the dataset by developing a learning model for a first subset of the N subsets.

The present invention relates to the technical field of water level prediction of ship nuclear power voltage regulators, and particularly relates to a voltage regulator water level prediction method based on a cost-sensitive LSTM cyclic neural network. The voltage regulator water level prediction method based on the cost-sensitive LSTM cyclic neural network comprises the following steps: S1, selecting p parameters (as shown in the specification) with a relatively high degree of coupling with a voltage regulator water level as input parameters; S2, constructing a LSTM-based voltage regulator water level prediction model and a framework thereof; and S3, using a BPTT algorithm to train and optimize the water level prediction model constructed in step S2. According to the method provided by the present invention, the LSTM model can better approximate the true value of the water level than the SVR model and the BP neural network model, and has stronger learning ability and prediction ability, and the cost-sensitive LSTM model has better precision and faster convergence.

The joint knowledge embedded method based on cost sensitive learning comprises the steps of S 1 establishing a training set consisting of a triple score function through a knowledge base, S 2 establishing a triple score function based on entities and relational embedded vectors, and establishing an optimization objective based on maximum margin under the condition of only considering context relation of the entity level. S 3 establishing cost sensitive joint embedded models. The joint knowledge embedded method based on the cost sensitive learning is characterized in that each entity and each relationship in the knowledge base and the knowledge mapping are respectively embedded into lower dimensional space on the basis of various related facts of the knowledge base. The layered contextual information of the knowledge base and the knowledge mapping is utilized better, making the embedded results satisfy the semantic structure of the knowledge base and the knowledge mapping better and enhancing predictive effects. The joint knowledge embedded method based on the cost sensitive learning has the advantages of expressing the visualization of the knowledge base and the knowledge mapping by utilizing the joint knowledge embedded method based on the cost sensitive learning, predicting the knowledge which is not within the knowledge base in answering question system.

The invention relates to a big data processing technology in intelligent information processing of a computer and particularly relates to an intelligent analysis early warning method for a dangerousness tendency of prison persons serving sentences. The intelligent analysis early warning method comprises the following steps: carrying out efficient collection and extraction on information of persons serving sentences in a prison information system to obtain basic data of behavior characteristics of the persons serving sentences; carrying out analysis judgment and early warning on the dangerousness tendency of the persons serving sentences by adopting a cost sensitive multi-stage semi-monitoring analyzing method. The intelligent analysis early warning method has the beneficial effects that 1, a lot of monitoring data of the persons serving sentences, which are provided by an existing system, are sufficiently utilized to automatically find potential behavior characteristics and behavior models of the abnormal persons serving sentences; the construction cost of the system is smaller and the feasibility is strong; 2 the method considers a cost problem of incorrect judgment of an early warning system and the influences caused by the early warning incorrect judgment are reduced to the greatest extent so as to meet actual requirements; 3 the method provided by the invention has a stronger self-adaptive ability; the early warning accuracy of the established early warning system is higher and the influences caused by the incorrect judgment are small.

The invention provides a malicious encrypted traffic detection method based on a logistic regression enhancement model, which is used for improving recall ratio on the premise of ensuring malicious encrypted traffic detection precision ratio. The method comprises the following steps: acquiring a training sample set S' and a test sample set X '; constructing a classifier set H (x) based on logisticregression; training the classifier set H (x) based on logistic regression; and obtaining a detection result of the malicious encrypted traffic. According to the method, the malicious encrypted traffic existing in the encrypted network traffic is distinguished by optimizing a cost function and adopting an iterative enhancement model of a plurality of cost sensitive classifiers. The method is usedfor solving the technical problem that in the prior art, due to the fact that abnormal encrypted traffic data is far smaller than benign encrypted traffic, the recall ratio is difficult to improve onthe premise that the precision ratio is guaranteed.

The invention discloses a software defect prediction method. According to the method, a sample with a class label and a sample without a class label are processed together, semi-supervised learning is utilized in the laplacian eigenmaps (LE), the LE method is improved, and meanwhile, in order to avoid that samples different in classification are mapped to a small low-dimensional neighbourhood, especially that defect samples are mapped to a defect-free sample neighbourhood, cost-sensitive information is introduced when the sample point distance is calculated through an LE algorithm. On the basis, the LE mapping precision is improved, and the discriminating performance of discriminating performance can be effectively improved through the method. The invention further provides a software defect prediction system. When the software defect prediction method and system are applied to an NASA database, it is proved through experiments that the effectiveness of the method is improved, and compared with other comparative methods, the classification performance of the method is improved to a certain extent.

The invention provides a cost-sensitive incremental face recognition method based on information entropy selection. The cost-sensitive incremental face recognition method is composed of a deep convolutional neural network part, an information entropy-based sample selection part and a cost-sensitive sequential three-branch decision classification part. The information entropy is used for assessingthe information amount of the classification result of the face recognition sample, so that the system can automatically assess the unlabeled sample information amount, and samples with large information amount are selected for manual labeling; a face recognition problem is regarded as a sequential process of information granularity from coarse to fine by utilizing the thought of three sequentialdecisions sensitive to cost, each iterative loop added with a marked sample is used as a decision step of the three sequential decisions, and the minimum cost recognition effect of the sample in eachdecision step is given according to the minimum Bayesian risk principle.

The invention discloses a biological information recognition method based on dynamic sample selection integration, mainly solving the problem of low correct recognition rate of subclass samples caused by data imbalance. The realizing process for solving the problem comprises the following steps: (1) a training set is divided into a series of balanced sub data sets by adopting a training set dividing method; (2) the obtained balanced sub data sets are divided into respective matrix classifiers as initial training sets; (3) on the matrix classifiers, cyclic training is carried out by adopting a dynamic sample selecting method; (4) a testing set is tested by decision functions obtained in each training, thus obtaining decision results; (5) weight of the decision results is calculated by adopting a cost-sensitive idea; and (6) the decision results of each time are weighted and integrated, thus obtaining the final recognition result. Compared with the prior art, the method has the advantages of high accuracy and low calculation complexity, the size relation between a correct ratio and a recall ratio can be regulated as required, and the method is used for recognizing biological information, network intrusion and financial fraud and detecting anti-spam.

The invention discloses a high-precision face recognition method for a complex face recognition access control system, belonging to the field of face image processing. The method comprises the following steps: face vector normalization, cost-sensitive face feature dimension reduction and model prediction. The method adopts the cost-sensitive face feature dimension reduction to give different misclassification cost for the different kinds of misclassification in the access control system. The extracted features meet the smallest bayesian risk criteria, the discriminant ability of the extracted features is increased, and further the accuracy of recognition is increased. In addition, the cost-sensitive face feature dimension reduction has robustness to the noises existing in the access control system, imbalanced categories and different kinds of data distribution density, and the method has a higher use value.

The invention discloses a fingerprint identification method based on depth hierarchy, which relates to the technical field of fingerprint identification and comprises a training part and an identification part. In the training part, according to the integrity and clarity of the fingerprint image, the fingerprint image is divided into low-quality fingerprint image and high-quality fingerprint image, and the fingerprint image is classified and marked. Using Resnet as the basic network, the quality evaluation network is constructed after many times of training. GAN technology converts low-qualityfingerprint image into high-quality fingerprint image, constructing quality enhancement network aft many times of training, finally, analyzing all high-quality fingerprint image, especially low-quality fingerprint image into high-quality fingerprint image, constructing cost-sensitive network aft many times of training; in the part of identification, the high-precision fingerprint identification can be achieved by using the constructed quality evaluation network, quality enhancement network and cost-sensitive network. The invention also discloses a fingerprint identification device, which is combined with a fingerprint identification method to improve the fingerprint identification accuracy.

The invention discloses a rehospitalization risk predicting method based on a cost-sensitive integrated learning model. The method comprises the following specific steps of: 1), acquiring medical andexternal environment data information, and constructing a multi-source high-dimension characteristic matrix; 2), performing high-dimension characteristic matrix nonlinear compression expression basedon an automatic encoder; 3), constructing an integrated learning model in which a cost-sensitive support vector machine is used as a weak learner; and 4), through characteristic processing of the step1 and the step 2, inputting a predicting set into a training model, and obtaining a rehospitalization risk predicting result. The method aims at patient demography information, previous hospitalization history, family history and an external environment characteristic and constructs the multi-source high-dimension characteristic matrix, thereby extracting more characteristic information which fully reflects the health condition of the patient. Based on high-dimension characteristic matrix nonlinear compression expression of the automatic encoder, dimension reduction on a sparse characteristicis realized. For aiming at a sample disproportion problem, the integrated learning model in which the cost-sensitive support vector machine is used as the weak learner is constructed, thereby improving rehospitalization risk identification precision.

The invention discloses a cost-sensitive support vector machine locomotive wheel detecting system and a detecting method thereof. The system comprises a data preprocessing module, a cost-sensitive support vector machine training module, a parameter optimizing module, an optimal cost-sensitive support vector machine classifying module, a determining module and a wheel state output module. The detecting method comprises eight steps totally. The parameter optimizing step is an adaptive mutation particle swarm optimization algorithm and has advantages of high robustness, global searching characteristic, etc. A space which continuously reduces in iteration is enlarged, and searching is performed in a larger space. Furthermore high population diversity is kept. Additionally, a possibility of finding out an optimal value by the algorithm is improved.

The invention relates to a multi-situational data and cost-sensitive integrated model-based place personalized semantic identification method. The method is specifically implemented by the following steps of 1) extracting effective features from various situational data of use logs of a smart phone, discovering user activities in acceleration data through clustering, and establishing user activity features of high-situational-level places; 2) according to activity distribution of the places, calculating semantic similarity of the places to obtain a cost matrix; 3) performing modeling on the features of the places in combination with the cost matrix, and introducing label-free place data for performing semi-supervised learning to obtain a plurality of cost-sensitive base classifiers; and 4) integrating the base classifiers to output an identification model, and performing personalized semantic identification on the places accessed by users. According to the method, the personalized semantic identification of the places is performed in combination with situational perception, cost-sensitive learning and semi-supervised learning; and the method has a wide application prospect in the fields of pervasive computing, location-based services and the like.

The invention discloses an average error classification cost minimized classifier integrating method. The method comprises the following steps of: 1, acquiring a training sample set; 2, initializing a sample weight and assigning an initial value; 3, iterating for T times, and training to obtain T optimal weak classifiers, wherein the step 3 comprises the following sub-steps of: 31, training weak classifiers on the basis of the training sample set S with the weight; 32, regulating the sample weight according to the results of the step 31; 33, judging whether t is smaller than T, if so, making t equal to (t+1) and returning to the step 31, otherwise, entering a step 4; and 4, combining the T optimal weak classifiers to obtain the optimal combined classifier. Compared with the prior art, themethod has the advantages that: classification results can be gathered in a class with low error classification cost in real sense, and on the premise of not requiring the classifiers to be independent of one another directly, the training error rate is reduced along with the increase of the number of the trained classifiers and the problem that the classification results can be only gathered in a class with the lowest total error classification cost in the conventional cost-sensitive learning method is solved.

A multi-agent system is provided for automatically acquiring desired information from one or more information sources. The multi-agent system includes a plurality of data provider filter agents associated with the one or more information sources. The data provider filter agents are configured to search for the desired information within the respective information sources based on an assessment of the one or more information sources. The multi-agent system also includes a content extraction agent configured to acquire a plurality of articles containing the desired information from the one or more information sources based on the search.