121 results about "Real time learning" patented technology

A system providing a mobile foreign language database and associated equipment. The system enables real time or near real time searching for foreign language business and real time or near real time learning of foreign languages.

The invention provides a specialized course recommendation method and system suitable for a network education platform. The specialized course recommendation method and system is different from a universal information recommendation method and system or information recommendation methods and systems in other fields. A tree data structure is formed for massive course resources in the network education platform, and a professional knowledge system and course learning characteristics of massive students are comprehensively considered, so that a user portrait model is formed; and based on this, classification mode identification is realized by utilizing learning histories and learning states of individual students, and then personalized course recommendation is generated from a big data course resource library according to categories which the students belong to and course correlativity.

Generalized learning rules may be implemented. A framework may be used to enable adaptive signal processing system to flexibly combine different learning rules (supervised, unsupervised, reinforcement learning) with different methods (online or batch learning). The generalized learning framework may employ average performance function as the learning measure thereby enabling modular architecture where learning tasks are separated from control tasks, so that changes in one of the modules do not necessitate changes within the other. Separation of learning tasks from the control tasks implementations may allow dynamic reconfiguration of the learning block in response to a task change or learning method change in real time. The generalized learning apparatus may be capable of implementing several learning rules concurrently based on the desired control application and without requiring users to explicitly identify the required learning rule composition for that application.

The invention relates to a motor rotating-speed tracking control method based on a self-adaptive fuzzy neural network, wherein rotating-speed and current double closed-loop control is adopted, an outer ring is a rotating-speed ring, a sliding-mode control theory-based fuzzy neural-network controller (SMFNN) is designed, an inner ring is a current ring, and a PI (Proportional-Integral) controller is adopted; a fuzzy neural-network rotating-speed controller comprises two parts, wherein one part is a PID (Proportional-Integral-Derivative) controller, and the other part is the fuzzy neural network, online real-time learning is carried out through the fuzzy neural network by utilizing a parameter correcting method designed on the basis of the sliding-mode control theory, and the two parts jointly act to obtain the output ir of the rotating-speed controller, i.e. a difference obtained by subtracting the output iFNN of the fuzzy neural network from the output iPID of the PID controller is used as the output ir of the rotating-speed controller. The control precision and the anti-interference performance of a motor speed-adjusting system can be improved through the control strategy of the motor rotating-speed tracking control method.

A learning fatigue recognition interference method based on facial expression recognition comprises the following steps: face region detection; facial feature extraction, comprising the establishment of a face expression facial model; expression classification recognition, namely carrying out classification recognition based on the face expression facial model; carrying out fatigue interference when recognizing fatigue expression or mental fatigue, and returning to continue to catch network learner picture image for next recognition after presetting shorter time; returning to continue to catch network learner picture image for next recognition after suspending presetting longer time in a manner that a system enters suspension when determining the learner is in neutral or absorbed expression or no mental fatigue; returning to continue to catch network learner picture image for next recognition when returning error messages. The technical scheme provided by the invention can rapidly recognize the expression of network learners and realize real-time learning fatigue interference.

The invention discloses a real-time learning debutanizer soft measurement modeling method on the basis of Gaussian mixture models (GMM). The real-time learning debutanizer soft measurement modeling method includes training process Gaussian mixture models to acquire various Gaussian component parameters and building corresponding sub-models; computing posterior probabilities of to-be-predicted samples and local Mahalanobis distances of various Gaussian components by a Bayesian process so as to obtain weighted sample similarity definition indexes; reasonably selecting similar samples by the aid of the new similarity indexes for local modeling. The posterior probabilities indicate whether the to-be-predicted samples belong to the various Gaussian components or not. The real-time learning debutanizer soft measurement modeling method has the advantages that problems of process non-Gaussianity and nonlinearity can be effectively solved, characteristics of the to-be-predicted samples can be sufficiently extracted, the similar samples can be reasonably selected for real-time learning modeling, and accordingly the real-time learning debutanizer soft measurement modeling method is favorable for improving the model prediction precision.

In one exemplary embodiment, a method for a real-time learning and collaboration environment comprising a first user's computing device and a second user's computing device to access a first digital whiteboard generated by the first user's computing device. A first modification is received from a first user to the first digital whiteboard. The first digital whiteboard is modified according to the first modification. A second user's computing device is enabled to generate a second digital whiteboard. The second digital whiteboard can overlay a view of the first digital whiteboard. A second modification is received from a second user to the second digital whiteboard. The view of the second digital whiteboard overlaying the first digital whiteboard is modified according to the second modification. A master view of the first view of the first digital whiteboard and the second view of the second digital whiteboard is provided to a whiteboard management dashboard application in the first user's computing device.

The present invention provides mobile terminal having anti-theft function. It contains anti-theft module; positioning module; and alarm module. The present invention also provides mobile terminal anti-theft method, first of all judging handset whether illegally used, if illegal use then automation automatically triggering positioning process to obtain handset positional information; secondly transmitting handset positional information to alarm number. Above-mentioned handset anti-theft device and method save tradition handset anti-theft method complexity process, real time learning handset current position, convenient handset positional tracking.

The invention discloses a self-learning wheel chair control method based on change of a gravity center of a human body, and belongs to the field of pattern recognition and intelligent systems. According to the self-learning wheel chair control method, a pressure sensor is installed between a wheel chair seat and a framework so as to collect force distribution under a sitting position of the human body, two-dimensional areal coordinates are calculated, and real-time data of the center of the gravity are stored in an embedded type computer; and algorithm optimization is conducted to the number of neurons in an output layer, network initial weight value, a network neighborhood radius adjusting rule and the like according to a basic learning process of a normal self-organizing feature map (SOFM) algorithm, and therefore operating complexity is reduced, calculating instantaneity of the algorithm in application is improved, and the purpose that algorithms are controlled to be different according to difference of people is achieved. By utilizing the improved SOFM algorithm, and in the process of driving habit learning, rate of convergence of an SOFM clustering algorithm and learning efficiency are greatly improved, instantaneity of the algorithm and accuracy of cluster are improved, the requirement of wheel chair real-time learning and controlling is met, and the problem that manual parameter adjustment is fussy due to difference of driving habits of users is solved.

A practical approach to real-time learning for synthetic characters grounded in the techniques of reinforcement learning and informed by insights from animal training. The approach simplifies the learning task for characters by (a) enabling them to take advantage of predictable regularities in their world, (b) allowing them to make maximal use of any supervisory signals, and (c) making them easy to train by humans. An autonomous animated dog is described that can be trained with a technique used to train real dogs called “clicker training.”

The invention belongs to the technical field of mobile network services, and relates to a parking space search platform based on mobile network services. The parking space search platform comprises an information transmitting end, a server end and an information receiving end. A server of the server end marks acquired information and quantifies the information as dimension information amount of limit learning, real-time learning prediction is conducted according to the quantified dimension information, and information acquisition and marking are mainly divided into three large categories. According to a search method, parking space information can be transmitted to a smart phone terminal of a car owner in time according to the actual environment, the car owner can acquire the parking space information in time, and therefore the inconvenience caused by looking for parking space can be reduced.

The optimal insulin to be delivered by an insulin infusion pump is determined by using a reinforcement learning algorithm aiming to the personalized glucose regulation. The algorithm optimizes the daily basal insulin rate and insulin: carbohydrate ratio for each patient, on the basis of his / her measured glucose profile. The proposed algorithm is able to learn in real-time patient-specific characteristics captured in the daily glucose profile and provide individualised insulin treatment. An automatic and personalised tuning method contributes in the optimisation of the algorithm's performance.

The present invention relates to a method for controlling motions of a robot using evolutionary computation, the method including constructing a database by collecting patterns of human motion, evolving the database using a genetic operator that is based upon PCA and dynamics-based optimization, and creating motion of a robot in real time using the evolved database. According to the present invention, with the evolved database, a robot may learn human motions and control optimized motions in real time.

The control apparatus for an internal combustion engine includes a throttle opening learning value calculation unit for calculating a throttle opening learning value based on a deviation between a target throttle opening and a learning throttle opening, controls the throttle opening by a learning corrected target throttle opening obtained by correcting the target throttle opening with the throttle opening learning value, and updates and stores a real-time learning value and a long time learning value based on a magnitude relation between values respectively obtained by adding the long time learning value to throttle openings respectively indicated by two effective opening area axis points of a correlation map, between which lies an actual effective opening area, and an actual throttle opening when the throttle opening learning value composed of the real-time learning value and the long time learning value is to be calculated.

The invention discloses a method for real-time monitoring of numerical control machining states of complicated structural components based on machining features. The method is characterized in that monitoring threshold values indicating whether the machining states are normal or not are established based on the machining features, and monitoring of the numerical control machining states is achieved according to the machining features; typical machining feature monitoring signals and a threshold value library are established through a cutting experiment, machining feature monitoring identifiers and machining feature monitoring threshold values in the matched threshold value library are added for numerical control machining programs before numerical control machining and serve as the basis for state monitoring in the machining process; for the machining features which do not exist in the threshold value library, machining feature samples are established in numerical control machining through real-time learning, and monitoring signals of new machining features arisen for the first time during machining serve as the basis for follow-up same-type machining feature state monitoring. The method can be not only suitable for monitoring of numerical control machining states during large-scale part production, can be also suitable for numerical control machining states during small-scale or even single part production and is wide in application range.

The invention belongs to the technical field of mobile network services, and relates to a parking space information sharing method based on smart phones. According to the parking space information sharing method based on the smart phones, vehicle owners carrying the smart phones serve as information users, and also serve as information collectors, the vehicle owners who serve as the information collectors send collected information to a server independently, the server marks the collected information, carries out quantification on the information to form dimensional information for extreme learning, and carries out real-time learning prediction according to the quantified dimensional information, meanwhile, the vehicle owners who serve as the information users obtain related parking space information which is processed through learning prediction from the server, and collection and marking of the information are mainly divided into three types. According to the parking space information sharing method based on the smart phones, the vehicle owners can timely obtain parking space information, inconvenience caused by looking for parking spaces is reduced, and the time for looking for the parking spaces is reduced.

The invention relates to the technical field of information and education, and relates to a real-time learning condition analysis system matching the classroom knowledge content and a real-time learning condition analysis method thereof. The real-time learning condition analysis system comprises a knowledge map generation module, a real-time learning condition data generation module and an analysis and comparison module. The knowledge map generation module generates an actual knowledge map according to the actual classroom teaching content. The real-time learning condition data generation module is used for performing algorithm analysis on the learning concentration situation of the onsite learning personnel so as to generate a concentration distribution diagram. The analysis and comparison module is connected with the knowledge map generation module and the real-time learning condition data generation module and establishes the correlation analysis view of the actual knowledge map andthe concentration distribution diagram with the time acting as the index so as to form the classroom knowledge content concentration distribution result. According to the real-time learning conditionanalysis system matching the classroom knowledge content and the real-time learning condition analysis method thereof, the knowledge map of the teaching content knowledge points is formed, and the knowledge points and concentration analysis are correlated with the time value acting as the index so as to form the classroom knowledge content concentration distribution result.

The embodiments of the invention disclose a method and apparatus for settling earning distribution. The method includes: based on the current state of an order and detailed information of the order, determining a plurality of earning distribution accounts; acquiring a settlement policy corresponding to the current state of the order; based on the settlement policy, keeping separate account of the earning distribution corresponding to the plurality of earning distribution accounts. According to the embodiments of the invention, a service platform can, based on the received detailed information of the order and the processing state of the current order, determine the amount of money to respective earning distribution account so as to, based on the state of the order, real-time distribute earnings to earning distribution accounts, help users of different earning distribution accounts to real-time learn the progress of earning distribution, and increase the real-timeness of earning distribution.

The invention provides a real-time studying system and method of wireless multimedia, referring to real-time embedded type stream media technique based on wireless network. The system is composed of a main control server, a stream media server, a wireless network AP meeting point and a mobile studying terminal. The two servers are connected with each other by a wired Ethernet. The mobile studying terminal is interlinked with the said two servers using wireless network. The main control server manages the user logging validation and the system alternation control. The stream media server maintains the study source and data and collects the data to produce the real-time stream media data issuance. The wireless network AP meeting point provides communication function of the mobile studying terminal and the said server. The mobile studying terminal is and portable embedded device and is interlinked with the said two servers using wireless network to alternate with the stream media. The invention has features of simple operation interface, low cost, convenient disposal, large capacity, rich interface, applicable for a great variety of application, in order to perform individuation study. The system can be used in university teaching, cooperation teaching and referring and so on.

A data processing apparatus processes input data and outputs the processed data. The data processing apparatus includes a data processing section and a real-time learning section. The data processing section processes the input data by a predetermined processing method and outputs the processed data. The real-time learning section controls such that the processing method is learned in real time and the data processing section processes the input data by the learned processing method, so that the output data is improved as time elapses.

Disclosed is a system, method, and computer program product for relating a metadata tag to a television program identifier and associating a hashtag with metadata associated with the television program. A mapping server receives a request that includes a metadata tag (e.g., a hashtag) for a social object, such as a comment or a post to a social network. The mapping server generates a graph to determine a set of television program identifiers likely associated with the metadata tag. Furthermore, the mapping server compresses graph data over time to reduce the time needed to perform a query to the graph.

The invention provides a real-time learning-based auxiliary word writing system and a real-time learning-based auxiliary word writing method. The real-time learning-based auxiliary word writing system comprises a paperwork characteristic analysis module for analyzing the information of the currently edited document to determine the paperwork type of the document, a real-time analysis module for analyzing the content of previously written document in real time to extract the content similar to the current edited document, and a content prediction and analysis module for uniformly processing the content acquired from the paperwork characteristic analysis module and the paperwork characteristic analysis module and sorting the acquired content to form prompt content. The invention also provides a real-time learning-based auxiliary word writing method. By the system and the method, the writing speed is greatly improved, the labor intensity is reduced, and the efficiency and the quality of the paperwork writing system are improved.

The invention provides a dynamic data comprehensive value evaluation method. The method includes constructing a data source module, constructing a retrieval module, retrieving error value evaluation,performing training and value evaluation on retrieval data, meanwhile comparing a final evaluation result and the previous plurality of evaluation initial values, and performing analysis by combiningerror values of retrieval and assessment modules to update the optimized error weight and continuously improve the accuracy of the evaluation result. A dynamic data comprehensive evaluation and analysis system of the method adopts the modular design for a data analysis system. By using the retrieval module, users can retrieve relevant target data directly and conveniently, use multi-thread multi-interface modules, improve calling speed and respond to dynamic real-time data access flexibly. The data processing accuracy is improved as much as possible by real-time learning and comprehensive adjustment of the data error weight.

The invention discloses an integrated real-time learning industrial process soft measurement modeling method based on multi-objective optimization, and belongs to the field of industrial process softmeasurement modeling. The invention aims to solve the common problems of redundancy and process nonlinearity of industrial data. An evolutionary multi-objective optimization method is adopted to optimize input variables and model structures in a historical sample database, variables irrelevant to quality variables or weakly relevant to the quality variables are removed, the sample quality of the database is improved, and meanwhile, the relation between the model complexity and the prediction precision is effectively balanced. Besides, a part of samples similar to the query samples are selectedfrom the optimized historical sample library to construct a local extreme learning machine model, and a selective integration strategy is adopted to integrate the Pareto optimal solution obtained through multi-objective optimization, so that the nonlinear problem of the industrial process can be effectively solved. By optimizing the modeling data structure and the extreme learning machine model structure, the prediction precision and the calculation efficiency of industrial process soft measurement modeling are improved.

In a pressure control system having a solenoid-operated fluid valve that has an output hydraulic pressure which varies in accordance with a solenoid input signal, a dynamic learning block is configured to adjust the initial, default values for control points stored in a pressure-current (P-I) data table based on observed (measured) operating points that reflect the solenoid's actual transfer characteristic. A feed forward control block is configured to generate the solenoid input signal having a level based on the adjusted control points in the data table, which improves the accuracy of the solenoid input signal. An adjustment method uses a plurality of circular buffers each configured to store observed operating points falling within a respective range, and provides a mechanism to allow adjustment of the control points based on only partial data.

A self-learning system of providing guidance on using inhaler devices, provided using a portable or mobile computing device. Guidance comprises of multiple stages(modes), and provides a unique feature of real-time learning using a camera device. It provides simultaneous viewing of the instructional video and the patient's real-time video in real time. The said guidance is provided in step-by-step instructions in an incremental manner of increasing complexity. The steps of the guidance are linked together by an audio-visual indicator. The self-learning system of using inhaler devices has various features including action-focusing videos, a technique to superimpose an animation on the learner's real-time video and virtual placebo inhaler. The said system of using inhalers also provides guidance on learning, practicing / reinforcing the correct method of using inhalers and assists the patient in using the inhaler on a regular basis by logging, analyzing and alerting about the time and dosage of the inhaler.

The invention relates to the field of distance education, and provides a remote user attention evaluation method based on multi-modal interaction. The method and the system aim to solve the problems that in remote education, the real-time learning state of a user is difficult to supervise, the learning quality cannot be evaluated, and the learning effect of the user is greatly affected. The methodcomprises the steps: acquiring multi-modal information of the user to be evaluated through an information acquisition module; judging whether the to-be-evaluated user is a target user or not according to the multi-modal information, and if not, sending prompt information; if the to-be-evaluated user is the target user, sending instruction information to the to-be-evaluated user, and receiving response information of the to-be-evaluated user; and evaluating the state of the to-be-evaluated user according to the response information. Evaluation of the real-time learning state of the user in distance education is realized so as to supervise the learning of the user.

The invention discloses an online reinforcement learning transaction system and method. The system comprises a collection module used for collecting transaction data of a transaction market, a feature mapping module used for performing continuous space mapping on the transaction data to obtain a reinforcement learning state, a reinforcement learning model used for performing Q function approximation in an iterative mode and updating model parameters, an output module used for outputting a Q value of each operation according to the reinforcement learning model and a reinforcement learning state obtained after each iteration and extracting a maximum value in Q values of all operations, and a calculation module used for evaluating a transaction effect according to the maximum value in the Q values of all the operations. According to the system, complex signals of the transaction market can be learned, and the market change can be learned in real time and adapted, so that the practicality and reliability of the system are improved; and the system is simple and easy to realize.