490 results about "Decision function" patented technology

This invention provides a system, method and apparatus for facilitating handoffs from a first communication network to a second communication network, the first communication network and second communication network being heterogeneous with respect to each other. The system, method and apparatus may further include a contextual information server, which stores contextual elements corresponding to a user device and the operating environment of the user device, and a handoff decision function module that evaluates at least one of the contextual elements to determine whether to handoff user device communications from the first communication network to the second communication network. The method and apparatus may further include obtaining at least one contextual element corresponding to a user device and the operating environment of the user device, evaluating the at least one contextual element with a handoff decision function module to establish a handoff decision, establishing a handoff decision, and notifying the user device of the handoff decision. The method for facilitating handoffs from a first communication network to a second communication network may further include receiving a received signal strength indication, receiving a link quality determination, receiving a characteristic of the user device, and determining the location of the user device.

A bandwidth efficient advanced modulation waveform modem using concatenated iterative turbo coding and continuous phase modulation is disclosed. A demodulator in the modem has a turbo decoder and a decision feedback carrier and time tracking algorithm to track a carrier and adjust timing. The decision feedback carrier and time tracking algorithm may use an APP decoder as a decision device to provide symbol decisions at a high error rate and low latency for a coded input data stream. A symbol phase estimator produces a symbol phase error estimate from the symbol decisions. An erasure decision function decides which symbol decisions are correct and which symbol decisions are erasures. A carrier tracking function receives the symbol phase error estimates when the symbol decisions are correct and receives erasure inputs when the symbol decisions are erasures to maintain carrier tracking.

The invention discloses a method for quickly and accurately detecting and tracking a human face based on a video sequence, which relates to the technical field of mode identification. The method comprises the following steps of: 1, extracting a video frame image from a video stream; 2, preprocessing the video frame image, namely compensating light rays, extracting skin color areas, performing morphological processing and combining the areas; 3, detecting the human face, namely representing the human face by using Harr-like characteristics and detecting the human face by using a cascaded Adaboost algorithm with an assistant decision function; 4, establishing the characteristics of the human face, namely detecting the area characteristics of the detected human face and the shape characteristics of the edge profile of the human face; 5, tracking the human face, particularly tracking the human face by using a human face area characteristic model when an intersection does not occur in a human face area, and further matching when the intersection occurs according to the shape characteristics of the edge profile of the human face; and 6, extracting the sequence of a human face image. By the technical method, the human face can be detected and tracked quickly and accurately on the basis of the video sequence.

A Session Admission Control (SAC) for negotiating admission control in a multi-services communications network including multicast services is described. The module distributes the admission process between a centralized decision function (SAC-PDP) and a distributed decision function (SAC-M) in a fashion that solves admission control scaling problems. The mechanism for interaction between the SAC-PDP and SAC-M is defined. Mechanisms are defined for the SAC-PDP to discover or learn the network capacity against which the admission control decisions will be made. Systems are also described for incorporating SAC-M in multicast replication points in the network, allowing multicast replication points to participate in the admission control process.

The invention provides an image quality evaluating method based on a support vector machine. The method comprises the following steps: first, a preprocessed image sample is selected and extracted according to characteristic value, a processed sample set is respectively divided into a training set and a testing set; secondly, the training set is used for training the support vector machine, the number of the support vector machine is ensured according to a certain level which is required by a system, thus ensuring each support vector machine to be trained, wherein, an input sample is the characteristic value of the image and an output sample is the level of the image quality; thirdly, the trained support vector machine is used for adjusting and optimizing correlation parameters with the testing set and determining the parameter of the decision function of the optimal hyperplane of the support vector machine model; and finally, the support vector machine model which is trained and optimized is used for evaluating the quality level of the image sample. The invention has the advantages of little required sample, fast arithmetic speed, high precision, good performance, strong popularization, etc.

Kernels (206) for use in learning machines, such as support vector machines, and methods are provided for selection and construction of such kernels are controlled by the nature of the data to be analyzed (203). In particular, data which may possess characteristics such as structure, for example DNA sequences, documents; graphs, signals, such as ECG signals and microarray expression profiles; spectra; images; spatio-temporal data; and relational data, and which may possess invariances or noise components that can interfere with the ability to accurately extract the desired information. Where structured datasets are analyzed, locational kernels are defined to provide measures of similarity among data points (210). The locational kernels are then combined to generate the decision function, or kernel. Where invariance transformations or noise is present, tangent vectors are defined to identify relationships between the invariance or noise and the data points (222). A covariance matrix is formed using the tangent vectors, then used in generation of the kernel.

The invention relates to a remote maintenance decision system of engineering machinery and a method thereof. The system comprises an information acquisition terminal, a client and a maintenance decision center. The information acquisition terminal acquires characteristic parameter information of a part of the engineering machinery in real time. The maintenance decision center comprises a life cycle phase determination module, a part residual life prediction module and a fault diagnosis and maintenance decision function module. The life cycle phase determination module receives the real-time characteristic parameter information and determines a present life cycle phase of the part. The part residual life prediction module carries out residual life prediction on the part at a normal phase or a performance decline phase. The fault diagnosis and maintenance decision function module carries out fault reason diagnosis on the part at a failure phase and provides a maintenance scheme. The client displays a residual life, or a fault reason and the maintenance scheme of the part. According to the system and the method of the invention, the life prediction or the fault diagnosis can be carried out on the part, and remote maintenance is carried out on the engineering machinery.

The invention discloses an unmanned aerial vehicle autonomous air combat decision framework and method, and belongs to the field of computer simulation. The framework comprises an air combat decisionmodule, a deep network learning module, an enhanced learning module and an air combat simulation environment which are based on domain knowledge. The air combat decision module generates an air combattraining data set and outputs the air combat training data set to the deep network learning module, and a depth network, a Q value fitting function and a motion selection function are obtained through learning and output to the enhanced learning module; the air combat simulation environment uses the learned air combat decision function to carry out a self-air combat process, and records air combat process data to form an enhanced learning training set; the enhanced learning module is used for optimizing and improving the Q value fitting function by utilizing the enhanced learning training set, and an air combat strategy with better performance is obtained. According to the framework, a Q function which is complex in nature can be more accurately and quickly fitted, the learning effect isimproved, the Q function is prevented from being converged to the local optimum value to the largest extent, an air combat decision optimization closed-loop process is constructed, and external intervention is not needed.

Methods for detecting the presence or absence of ice or liquid water on surfaces are provided. In a first embodiment of the invention, a reflectance spectrum from a surface to be tested is measured using any suitable near-infrared optical system. The midpoint wavelength of the transition or step of the measured reflectance spectrum near 1.4 microns is calculated, and compared to a decision threshold wavelength. If the midpoint wavelength is less than the decision threshold wavelength, the presence of liquid water on the surface is indicated. If the midpoint wavelength is greater than the decision threshold wavelength, the presence of ice is indicated on the surface. In a second embodiment of the invention, at least three reflectance levels are measured at three wavelengths, and a dimensionless decision function is applied to the measured reflectance levels. Output of the decision function is compared to at least two pre-defined ranges. If the output is within a first range, the presence of liquid water on the surface is indicated. If the output is within a second range, the presence of ice on the surface is indicated. If the output is between the pre-defined ranges, the absence of ice or water on the surface is indicated.

The invention discloses an identification method of fine-grained attributes of pedestrians under complex scenes. A classification model is adopted for fine-grained attribute identification on sub-components of a detected pedestrian; association analysis is respectively carried out on identified attributes and a pedestrian gender, and attributes with high correlation are selected for multi-task learning; then convolutional neural network (CNN) models constructed by multi-task learning are trained, and results of the convolutional neural network models with a highest identification rate are selected for the multiple attributes to use the same as a final result; and finally, a gender attribute of the pedestrian is judged according to a customized decision function. The method can realize overall-to-local detection of the pedestrians in the complex scenes, realize more accurate detection and identification of the attributes of pedestrian sub-components, and avoid interference of backgrounds and other information, also solves, at the same time, the problem that detection correctness rates of models on small objects are low, and has higher information accuracy.

The invention discloses a power distribution network fault processing method based on multi-source information fusion. The method includes five steps of 1) multi-source information searching, 2) fault data information collection, 3) fault model matching, 4) fault location and 5) fault processing. According to the method, a fault source is located by means of multi-source information collection and information matching, then fault processing and disclosing are conducted, the specific fault source can be accurately and timely located when operation faults of a power distribution network occur, corresponding processing is conducted, an auxiliary decision making function is provided for first-aid repairing and management of the faults, automation building of the power distribution network can be promoted quickly, and good application prospects are obtained.

Data processing apparatus is provided for evaluating answers to respective query items considered to be represented by respective points within a region of feature space, and the apparatus comprises an input (10) which receives such a query item. The region is subdivided into subregions according to at least first and second subdivisions. A subregion identifying portion (20) identifies, for each such subdivision of the region, which subregion of the subdivision contains the point representing the received query item. A partial answer retrieval portion (30) has access when the apparatus is in use to a store (40) of precalculated partial answers for at least some the subregions of the subdivisions, and retrieves from the store the partial answers for the or each identified subregion that is present in the store. An answer calculation portion (50) calculates an answer to the received query item based on the retrieved partial answers, and an output (60) outputs the calculated answer.

The invention discloses a SAR target recognition method based on a sparse least squares support vector machine, which belongs to the technical field of image processing and mainly solves the problem that the existing method need a long time for SAR target recognition. The realization process comprises the following steps of: firstly respectively implementing feature extraction to the selected target images with known classification information and images to be recognized to obtain training samples and test samples; and then applying iterative training to the training samples by using the combination of incremental learning method and reversal learning method to select a sparse support vector set and obtain a Lagrange multiplier and deflection corresponding to the support vectors in the set; and finally using a classification decision function to recognize the test samples according to the obtained support vector set, the Lagrange multiplier and deflection corresponding to the support vectors. The invention has the advantage of shortening recognition time under the condition of equivalent recognition precision and can be used for detection and recognition of SAR target.

The invention discloses a weighted decision and random scheduling method based on physical heterogeneous redundancy. Different unicast, multicast routing protocols are executed in each routing protocol processing unit, and a calculation result is output to a multi-mode decision unit; the multi-mode decision unit decides the calculation result of each routing protocol processing unit, and deliversand outputs a final decision result to a data forwarding plane; a redundancy scheduling unit builds a routing protocol pool, performs protocol setting on each routing protocol processing unit, recordsand calculates each decision result, and then dynamically and randomly schedules the routing processing unit according to the calculation result. Due to design of the routing mechanism with the heterogeneous redundant decision function, the invention provides a weighted multi-mode decision method to compare output results of multiple heterogeneous routing function execution bodies, so as to perform multi-mode decision on final routing selection, dynamic random scheduling selection is performed on the execution bodies through a trust degree weight, then a simulation defense capability of a switch routing control plane is realized.

The invention discloses a computer aided model construction method based on deep learning gastric cancer pathological sections, and belongs to the technical field of artificial intelligence. The method uses a 121-layer dense-connected convolutional neural network to perform image recognition. A dense block structure in a DenseNet allows the high-level part of the network to acquire shallow features, which greatly reduces the over-fitting phenomenon. At the same time, the model has a large number of layers, which can fit more complex and smoother decision functions. Although the number of layers is large, the number of parameters of the model is not large, which saves resource consumption. In order to further avoid over-fitting, a training mechanism for migration learning is adopted. The model will be pre-trained on an ImageNet dataset to give the model a strong image feature extraction capability. The main optimization of the model during formal training can be better focused on how toextract the features of the diseased area, and the utilization rate of the data is greatly improved.

A wireless communication system that implements cell reselections selects a timing of a cell reselection in association with a mobile station (MS) by transferring, by a layer at least as high as Layer 3 to a cell reselection decision function residing in a layer lower than Layer 3, information indicating whether a present time is a more favorable time or a less favorable time for a cell reselection and initiating, by the cell reselection decision function, a cell reselection at a time based on the information received from the layer at least as high as Layer 3. In another embodiment of the invention, the transferred information may instead, or in addition, indicate a future time at which to perform a cell reselection. By selectively timing a cell reselection, an impact of a cell reselection on an application being executed by the MS is minimized.

The invention discloses a wireless sensor network fault diagnosis method, which comprises the following steps of: acquiring link quality signals of a wireless sensor network in a normal state and a fault state; performing wavelet packet decomposition and coefficient reconstruction by using a wavelet transform method, and extracting a characteristic vector of a fault; constructing a coding matrix by adopting an error correcting output coding method to realize network fault multi-classification; constructing a decision function of multi-classification problems by using an LSSVC learning mechanism, and establishing a relationship model between the system state and the characteristic vector; and identifying, diagnosing and processing a generated fault, a potential fault type and an area according to a network running decision function value. The method has wide application prospect in small and medium-scale WSN application system.

The invention provides a WIFI automatic association method and an intelligent terminal. The WIFI automatic association method includes the steps that the intelligent terminal STA conducts a preliminary screening on a monitored AP collection according to a busy degree and signal strength of each wireless access point AP; STA calculates jamming intensity of each AP according to average bottom noise of a communication channel and the busy degree and the signal strength of each AP, and rechecks the AP collection after the preliminary screening according to the ratio of the signal strength to the jamming intensity; STA calculates the decision function value of each AP in the AP collection after checking according to an optimal AP decision function; STA selects the AP to be associated with according to the busy degree, the signal strength, a signal-to-jamming ratio and the decision function value of each AP after rounds of selection. According to the technical scheme, the WIFI automatic association method and the intelligent terminal comprehensively consider coverage, capacity, interference, switching calling loss and switching expense and conduct WIFI automatic association, and web surfing experience of a user in a WIFI network is improved.

The invention discloses a method and a device used for detecting similarity between source programs. The method includes the following steps: two source program files are respectively optimally decoded to generate two binary files; the two binary files are disassembled to generate two assembly code units; decision function is adopted to calculate the two assembly code units to define the similarity between two source programs. The method can eliminate trouble caused by advanced plagiarism means such as code redundancy, sentence splitting and constant displacement, etc. on a program semantic level by the execution of optimal decoding and disassembling.

A sound processing device includes a separation unit configured to separate at least a music signal and a speech signal from a recorded audio signal, a noise suppression unit, a music feature value estimation unit, a speech recognition unit, a noise-processing confidence calculation unit, a music feature value estimation confidence calculation unit, a speech recognition confidence calculation unit, and a control unit configured to calculate at least one behavioral decision function of a speech behavioral decision function associated with speech and a music behavioral decision function associated with music based on a noise-processing confidence value, a music feature value estimation confidence value, and a speech recognition confidence value and to determine behavior corresponding to the calculated behavioral decision function.

The invention discloses a model training method and an apparatus thereof. The method comprises the following steps of constructing a decision function model and training the decision function model so as to acquire an importance measurement value of a plurality of training samples; according to each importance measurement value, carrying out descending sort on each training sample, and making the sorted training samples generate a characteristic subspace; classifying each training sample in the characteristic subspace and acquiring a plurality of classification results, weighting and integrating the plurality of classification results so as to acquire a predicted score; and determining whether the predicted score is the same with a preset score, and when the predicted score is not the same with the preset score, adjusting a parameter of the decision function model. By using the method, accuracy of the predicted score is increased.

Methods and systems for modeling user arrival and choice in the context of off-street parking solutions. A first component models the arrival and duration of stay of users as a function of time, taking into account different user profiles (or “clusters”), captured by a latent variable. A second component provides a ranking function (for each user cluster), wherein the input features describing the “choice” constitute status variables associated different car park(s), and the output constitutes a preferred car park and a pricing scheme. The system simulates different user behaviors by assuming some standard groups of users will behave similarly. Groups of users or user profiles are learned automatically. The profiles are then employed as a key element for automatically learning a decision function of parking users, and automatically learning one decision function per profile.