50 results about "Score distribution" patented technology

A modular scoring system using rank aggregation merges search results into an ordered list of results using many different features of documents. The ranking functions of the present system can easily be customized to the needs of a particular corpus or collection of users such as an intranet. Rank aggregation is independent of the underlying score distributions between the different factors, and can be applied to merge any set of ranking functions. Rank aggregation holds the advantage of combining the influence of many different heuristic factors in a robust way to produce high-quality results for queries. The modular scoring system combines factors such as indegree, page ranking, URL length, proximity to the root server of an intranet, etc, to form a single ordering on web pages that closely obeys the individual orderings, but also mediates between the collective wisdom of individual heuristics.

The present invention relates to improving the identification rate for a person without lowering the rate of excluding another person probabilistically. The invention is a personal identification system that finds the verification results of feature data extracted form data showing the physical features of the individual who is the verification source against the pre-registered feature data of the verification object, and identifies the individual by whether or not the degree of similarity obtained by the verification results exceeds a separate threshold value set according to pre-registered feature data, and wherein a separate threshold value calculation device 23 is provided that calculates the separate threshold value based on the verification score distribution of feature data extracted from data showing the physical features of the individual who is the verification source and the corresponding feature data, and the verification score distribution of the feature data extracted from data showing the physical features of the individual who is the verification source and non-corresponding feature data.

A method of authentication is provided that includes capturing palm biometric data from an individual, transmitting the biometric data to an authentication system, and generating an authentication biometric template from the captured biometric data with the authentication system. Moreover, the method includes conducting a plurality of verification matching transactions between the authentication biometric template and enrollment biometric templates stored in the authentication system, and between the authentication biometric template and imposter biometric templates stored in the authentication system. Furthermore, the method includes updating a genuine matching score distribution with at least one genuine matching score, updating an imposter matching score distribution with at least one imposter matching score, and calculating a maximum genuine matching score and a minimum imposter matching score. Additionally, the method includes confirming an identity of the individual when the minimum imposter matching score is greater than the maximum genuine matching score.

The invention discloses a test question difficulty analysis method and system. The method comprises the following steps of: obtaining a to-be-analyzed test question; extracting question features of the to-be-analyzed test questions, wherein the question features comprise a question surface feature, an analysis feature and an answer feature; and obtaining a difficulty score of the to-be-analyzed test question according to the question features of the to-be-analyzed test question and a pre-constructed test question difficulty prediction model. According to the method provided by the invention, the difficulty score of the to-be-analyzed test question is predicted according to the question features of the to-be-analyzed test questions, but is not predicted according to history score distribution condition of the to-be-analyzed test questions, so that test question difficulty evaluation can be carried out on cold data conditions, namely, the questions which are not done by students.

The invention discloses a personalized assessment method for a subject grasping ability of a student. According to the method, a middle school student learning a standard subject system is mainly usedas the research object, personalized assessment of the subject grasping ability of the student specifically is assessment of the subject knowledge point grasping ability of the student. The method comprises the steps that a student information database is established; difficulty coefficients of questions are calculated according to a question answering score of the student; a single-time assessment result of the student in a current examination is calculated according to question information and the current score of the student; a personalized assessment result of the accumulated knowledge point grasping ability of the student is calculated based on historical scores of the student and historical question information; and student score distribution and a relative grade of the student arecalculated according to the historical scores of the student, and then a personalized assessment result of the knowledge point grasping ability of the student relative to same-grade students is calculated. The method has good generalization performance and adaptability, and the recognition effect has the advantages of objectivity, reliability and comprehensiveness.

The invention discloses a hierarchical classification method and a device. The method comprises the following steps: receiving input images via a neural network and identifying the images to output normalized scores; searching a maximum score in the normalized scores and determining the corresponding class; judging whether the maximum score is greater than a threshold or not; outputting the maximum score and the corresponding class as identification results when the maximum score is greater than the threshold; converting score distribution into rough class score distribution when the maximum score is not greater than the threshold, then circularly performing the steps of searching the maximum score in the normalized scores, determining the corresponding class and judging whether the maximum score is greater than a threshold or not, stopping circulating when determining that the maximum score is greater than the threshold, and outputting the maximum score and the corresponding class asthe identification results. Through the above manner, the hierarchical classification from fine to rough is achieved, so that the problem of wrong identification or missing identification caused whenall the fine classes of scores are low in the existing image identification task can be solved.

Systems and methods include: collecting digital threat scores of an incumbent digital threat machine learning model; identifying incumbent and successor digital threat score distributions; identifying quantiles data of the incumbent digital threat score distribution; collecting digital threat scores of a successor digital threat machine learning model; calibrating the digital threat scores of the successor digital threat score distribution based on the quantiles data of the incumbent digital threat score distribution and the incumbent digital threat score distribution; and in response to remapping the digital threat scores of the successor digital threat score distribution, publishing the successor digital scores in lieu of the incumbent digital threat scores based on requests for digital threat scores.

The embodiment of the invention provides an image evaluation method and device and a computer readable storage medium, and relates to the technical field of image processing. The image evaluation method comprises the steps of obtaining a to-be-evaluated image; calculating the to-be-evaluated image based on a preset image evaluation model to obtain score distribution information of the to-be-evaluated image; and calculating and obtaining an image evaluation value corresponding to the to-be-evaluated image according to the score distribution information. The accuracy of the image evaluation result can be effectively improved.

The invention provides an English oral automatic grading method and system. The method comprises the steps of: extracting multi-frame voice characteristics of voice signals to be detected; inputting the multi-frame voice characteristics into a depth nerve network DNN model, and extracting a frame posterior probability of each frame voice characteristic, wherein the DNN model is trained in advance, the input of the DNN model is the frame voice characteristics, and the output is the frame posterior probabilities of the frame voice characteristics to different pronunciations; carrying out distribution statistics on the frame posterior probability of each frame voice characteristic, and obtaining the global characteristic of the voice signals to be detected; and carrying out grading on the global characteristic by using a multilayer forward nerve network model MLP. Compared with a conventional grading method based on GMM, the method provided by the invention is high noise and channel robustness, higher discrimination is realized for pronunciation quality, and the distribution of obtained scores is more reasonable.

An image processing includes multiple pieces of image data capturing biological tissues, a scene determining device configured to determine a photographic scene based on the color moving image data, a score calculating device calculating a score indicative of seriousness of lesion of the biological tissues captured in the image represented by the image data, based on the image data, and a marking device applying marks indicative of a distribution of the scores on the image. The marking device executes a detailed marking process to apply the marks indicating the distribution of the scores in detail, and a simple marking process to apply the marks indicating the distribution of the scores in a manner simpler than the detailed marking process. The marking means executes one of the detailed marking process and the simple marking process in accordance with the result of determination of the photographic scene.

The invention provides a method for question bank quality evaluation. The method includes the steps that a question bank quality evaluation quantitative model is established, and quality evaluation indexes which comprise one or more of the question reliability index, the question validity index, the score distribution index, the question discrimination index and the question difficulty index are determined; a BP neural network is initialized according to the quality evaluation indexes; the initial weight value and the threshold value of the BP neural network are determined through a particle swarm IM-PSO algorithm based on an immune mechanism; a training sample composed of the quality evaluation indexes is used for training and obtaining the BP neural network for question quality evaluation according to the initial weight value and the threshold value. By the adoption of the method, the global searching capacity of the IM-PSO algorithm is adopted, due to the fact that the diversity of populations is enhanced, a local minimum can be prevented from being gotten into in the searching process, the searching accuracy and the operating efficiency of the algorithm are improved, and finally the purpose of network optimization is achieved.

The embodiment of the invention discloses a method and equipment for displaying body data. The method comprises the following steps: acquiring measurement data of a user; in accordance with three preset items, namely a health condition score distribution ratio, a computation rule and the measurement data, acquiring the body data of the user; and displaying the acquired body data of the user. According to the embodiment, the method disclosed by the invention comprises the steps of acquiring the measurement data of the user, acquiring the body data of the user in accordance with three preset items, namely the health condition score distribution ratio, the computation rule and the measurement data, and displaying the acquired body data of the user; since the displayed body data are calculated and acquired from three preset items, namely the health condition score distribution ratio, the computation rule and the measurement data, the body data can be specifically represented by health condition scores, so that the health status of the user can be directly and accurately displayed by a display result, and user experience is greatly improved.

The invention discloses an extraction method for technological entries representing achievements and capabilities of technological experts. The method includes the steps of firstly performing quantitative calculation on the technological achievements and calculation on the technological entries, and obtaining the technological entries and score distribution included in each achievement; then counting achievement scores on all the technological entries, and generating a technological entry tree graph with annotations, wherein the weights of vertexes are the achievement scores on the entries, and the weights of edges are combination coefficients, namely the ratio of subsequent vertex weights to precursor vertex weights; finally calculating a technological entry list directly included in theachievements of the experts according to the achievements of the experts and the technological entries contained in the achievements; expanding the technological entry list of the experts by using anassociated relationship among the entries in the technological entry tree graph; calculating achievement scores of the experts on the expanded entries by using the combination coefficients of the technological entry tree graph, and calculating contribution coefficients, namely the ratio of the achievements scores of the experts to the total achievement scores of the technological entries; selecting k technological entries with the highest contribution coefficients for representing main research areas and the achievements and capabilities of the experts.

The present invention provides a candidate voluntary filling simulation system and method. The candidate voluntary filling simulation system includes: a collection module, a voluntary filling module, a simulation analysis module and a recommendation module, wherein the collection module is used to collect control lines over the years, adjustments over the years The grade line, the distribution of the admission scores of each university over the years and the control line of the year; the voluntary reporting module is used to receive the externally input candidates' scores and voluntary information; the simulation analysis module is used to base the control line over the years, the grade adjustment line over the years, and each school The distribution of admission scores over the years, the control line of the year and the scores of candidates, analyze the admission rate of each school; the recommendation module is used to make recommendations for voluntary submissions based on voluntary information and admission rates. The scheme provided by the invention can effectively improve the acceptance rate of voluntary filling.

The invention discloses a speaker confirmation method adopting unsupervised clustering score regularity, which comprises the following steps of: testing a current speaker model of a registration set and a regularized data set containing L corpora to obtain L test scores; clustering the L test scores by using a K-means algorithm, and screening out a plurality of types of scores with the front meanvalues; adopting an expectation maximization algorithm to cluster a plurality of types of screened scores, introducing a Gaussian mixture model to fit score distribution for probability distribution,and after the Gaussian mixture model converges, taking a Gaussian unit mean value with the maximum mean value and a standard deviation as score regularization parameters of the current speaker model;regularizing L test scores by using score regularity parameters, and performing speaker confirmation by using the regularized scores.

The invention discloses a method and system for detecting an individual anxiety level, the method comprises the following steps: providing N attention suppression capability test signals for an subject, obtaining N attention suppression capability responses and N groups of eye movement data returned by the subject to the N attention suppression capability test signals; calculating the detection accuracy of the attention suppression capability of the subject according to the N attention suppression capability responses; according to the detection accuracy rate and the N groups of the eye movement data and the like, scoring the attention suppression capability of the subject to obtain a score of the attention suppression capability; inquiring a score distribution table of attention suppression capabilities corresponding to different anxiety levels; and predicting the anxiety level of the subject. According to the detection system, strict task-independent interference stimulation is adopted for detection, irrelevant information interference caused by a target is eliminated, the obtained detection result is stable and repeatable, and the problem that an oral reporting type scale lacksobjective and reliable detection indexes is avoided.

The invention discloses an intelligent test paper generation method and a device, computer equipment and a storage medium. The method comprises the following steps of: obtaining a pre-stored electronic document, crawing basic knowledge points from the pre-stored electronic document in a web crawler mode; and according to the basic knowledge points, generating basic test question bank, the problemof weak test question confidentiality caused by directly obtaining the existing test questions from the network is avoided; the confidentiality of the generated test paper is improved; meanwhile, whenan examination paper generation request is received, obtaining score distribution of knowledge points contained in the examination paper from the request; according to the score distribution and thecorresponding relationship between each basic test question and the score distribution of the knowledge points contained in the basic test questions, calculating the score distribution of the knowledge points; according to the method, selecting the basic test questions meeting the score distribution of the knowledge points contained in the examination paper from the basic test questions containedin the basic test question bank to serve as the target test questions, the test paper meeting the requirements is intelligently generated according to the distribution of the knowledge points, the test paper generation efficiency is improved, and the practicability of the generated test paper is higher.

The invention discloses a competitive sports teamwork mode analogy method based on chordal graph visualization. The method comprises the following steps: 1) establishing a foundation database of game related information; 2) constructing a chordal graph according to the assisting information of players in a single game live broadcast record, i.e. dividing each player into a scoring sector and an assisting sector, respectively representing a scoring record and an assisting frequency record of each player, connecting the corresponding assisting sectors and scoring sectors of both cooperated parties by a chord so as to obtain a detail record of a game; 3) clustering the players with a multidimensional scaling method according to the historical record data of mutual assisting information of players in a team to obtain the two-dimensional coordinates of m vectors so as to obtain the clustering information of m players. Compared with the prior art, the competitive sports teamwork mode analogy method has the characteristics that chordal graph visualization design is adopted so as to bring convenience for users to better know and analyze game score distribution and player cooperation situations.

An embodiment of the invention provides a virtual prize distribution method and device and electronic equipment and relates to the technical field of data processing. The virtual prize distribution method comprises steps as follows: scores of each player participating in a game task and completing the game task are acquired, and the scores are obtained according to contribution degree of the player to the game task; at least one prize corresponding to the game task is shown to the players participating in the game task; a score distribution result, aiming at the at least one prize, fed back byeach player is obtained, and the sum of the scores, distributed to the at least one prize, of each player participating in the game task is not higher than the scores of the player completing the game task; aiming at each prize, the prize is distributed to the player with the highest distribution scores for the prize. According to the virtual prize distribution method and device and the electronic equipment, reasonability of game prize distribution can be improved, and motivation of the players is improved.

The invention discloses an examination paper reading system and a marking implementation method thereof. The method comprises the following steps of: A, acquiring and displaying practical examination paper scanning data; B, searching and displaying examination point information corresponding to examination questions in the practical examination paper scanning data according to a corresponding relationship between the pre-stored examination questions and the examination point information thereof; and C, judging whether instructions of externally selecting the examination point information are acquired, if so, establishing and displaying a corresponding relationship between the selected examination point information and the practical examination paper scanning data. The system and the method can display the examination point information corresponding to each examination question of examinee paper and the score distribution condition corresponding to the examination point information so as to conveniently, efficiently and quickly count the mastering condition of each examination point in the examination paper and provide precise data support for comprehensively analyzing the whole examination or the answering condition of each examinee.