174results about How to "Reduce the influence of subjective factors" patented technology

The invention relates to a distributed buffer memory strategy adaptive switching method based on machine learning and a system thereof. The method comprise the following steps: carrying out evaluation on each buffer memory strategy based on a reference test, and confirming scene elements influencing buffer memory strategy performance appearance; collecting data set of each buffer memory strategy under different scene elements conditions; training the data set to obtain a buffer memory performance model; and deciding an optimal buffer memory strategy based on buffer memory performance model and on-line monitoring data in cluster environment, and executing buffer memory strategy switching when present buffer memory strategy is inconsistent with the optimal strategy. In the invention, machine learning method is employed to establish the buffer memory performance model which can accurately describes present scene, the model is updated periodically through constructing a performance data warehouse to improve the precision of the model further. Thus the distributed buffer memory strategy adaptive switching method in the invention is adapted to environmental dynamic change well to enhance flexibility and adaptability of buffer memory service.

The invention discloses a method for automatically identifying and classifying rock lithology in a deep learning mode, aiming at analyzing rock lithology in geological engineering. The method includesthe following steps: A. based on the kinds of rocks, acquiring rock images of different types, and grouping the rock images into a training group and a test group; B. taking a convolutional neural network Inception-v3 model as a pre-training model, acquiring image features by using a feature extraction model of the pre-training model; C. establishing a Softmax regression model; D. training a model for automatically identifying and classifying rock images; and E. testing the model for automatically identifying and classifying rock images. According to the invention, the method herein, by establishing the model for automatically identifying and classifying rock images, can analyze the geological conditions in an automatic and intelligent manner, greatly save labor and materials, and reducescost.

The invention relates to a track profile irregularity amplitude estimation method employing optimal belief rules based inference. According to the method, the mapping relationship between parametric variable input and yield output is modeled by a belief rule base. A corresponding change relationship between vibration frequency domain characteristic data of different measurement points and a track profile irregularity amplitude is described by building the belief rule base. By a sequence linear programming method, an initial belief rules based (BRB) model is optimized through limited historical data; and the effects on the model caused by subjective factors are reduced. According to the sequence linear programming (SLP) method, a nonlinear optimization problem of an original model is converted into a step-by-step linear optimization problem; and various parameters of the optimization model can be relatively simply and rapidly calculated, so that the track profile irregularity amplitude can be accurately and rapidly estimated through belief inference under the condition of given vibration frequency domain characteristic. According to the track profile irregularity amplitude estimation method, the estimation accuracy and the calculation efficiency of the model are improved; and the method has the advantage of being relatively efficient on a track profile irregularity system which needs to be monitored in real time.

The present invention discloses a work order data-based employee performance evaluation system and method. The system includes a standardized work order data adaptation interface, an intelligent performance data processing module, a performance scoring module and a performance display module; the standardized work order data adaptation interface is used for acquiring work order data so as to obtain standardized work order data; the intelligent performance data processing module is used for acquiring operation item work credits according to a pre-defined rule; the performance scoring module is used for a performance evaluation manager to carry out performance evaluation and scoring on performance appraisees so as to obtain performance evaluation scores; and the performance display module is used for carrying out visualization processing on the performance evaluation scores and displaying obtained processing results. With the employee performance evaluation method adopted, evaluation basis for performance evaluation is quantified, influence of the subjective factors on an evaluation result is decreased, and therefore, the work order data-based employee performance evaluation system and method can be fair, just, accurate and reliable.

The invention discloses fabric color fastness color difference grading method. It includes the following steps: copying fabric color character and transmitting to the computer to obtain homochromatic different spectra image in RGB color space coordinates; analyzing and measuring. It has the advantages of practical, reliable, easy operation, fast analyzing, automatic grading etc.

The invention discloses a coma degree evaluating method based on multiple indexes of non-linearity and complexity, which is used for monitoring the brain state of a patient in a coma, and can realize continuous coma depth monitoring, and coma phase grading and early warning. Multiple complexity indexes such as complexity, Lyapunov exponent, approximate entropy, related dimensions, and the like are extracted from coma electroencephalographic signals by adopting a nonlinear dynamics analytical method, and a comprehensive coma state exponent is obtained by combining a traditional GOS (Glasgow Outcome Scale) evaluation system and a traditional GCS (Glasgow Coma Scale) evaluation system; the coefficients of the correlation between the parameters are obtained by clinical experiments; a coma state grading database is founded and a parameter fusion coefficient is determined on the basis of clinical effects; and finally, complete conscious state grading evaluation exponents are founded to guide the treatment of a patient and make prognosis.

The invention relates to an assessing method for an influence of an office scene illuminating environment on vision health. The method comprises the following steps: measuring present working surface luminance, brightness, color temperature, lamp position, working environment area and illuminating power; adopting seven indexes, including vision comfort level, uniformity of luminance, strobe, brightness distribution, glare, color temperature and lighting power density; calculating the data of the seven indexes by repeatedly measuring the data of the practical illuminating scene; adopting an entropy weight method and confirming an index weight; and lastly, acquiring a comprehensive score. The invention provides the assessing method which is perfect in index system, easy in operation and capable of assessing the vision health for the office scene illuminating environment.

The invention discloses a method for monitoring mental diseases. The method comprises: acquiring data related to the mental disease condition of a patient; analyzing the acquired data; and sending out early warning information when the data meets a preset early warning condition. The invention also discloses a system for monitoring the mental diseases. The monitoring method can realize automatic monitoring of the mental disease condition by means of a computer system, and reduces the influences of doctor subjective factors; meanwhile, the monitoring method is convenient for patients to realize self management of the mental diseases and reduces the therapeutic burden of the mental diseases.

The invention discloses a method and device for detecting an abnormal user. The method comprises the steps that logs are acquired; the acquired logs are classified; the various classes are scored according to access parameters in the various classes, the classes with the scores ranking the first N are selected, and the logs in the N classes are taken as key logs; and group characteristics of the key logs are extracted according to a characteristic extraction algorithm, individual characteristics of a log of a target user are extracted according to the characteristic extraction algorithm, and whether or not the target user is an abnormal user is judged according to the similarity of the group characteristics and the individual characteristics. The device is independent of a real-time strategy system, has a longer time window and can obtain more accurate characteristics of the user.

The invention discloses a paper chip enzyme-linked immunoassay test card for a combined multi-tumor marker test. The test card is characterized in that upper, middle and lower filter paper chips, a double-sided tape and a paper-based slide frame are arranged between an upper shell and a lower shell; the upper, middle and lower filter paper chips refer to wax patterned filter paper chips which are provided with hydrophilic and hydrophobic regions and are alternately arranged respectively; a sample introduction and liquid feeding window is formed in an upper shell body; the upper filter paper chip is arranged between the upper shell and the double-sided tape; the middle filter paper chip is arranged between the double-sided tape and the lower filter paper chip, and sensitive regions which are distributed in arrays and used for fixing different reaction reagents are arranged on the middle filter paper chip; the water-based slide frame is arranged between the lower filter paper chip and the lower shell; the upper and middle filter paper chips are bonded through the double-sided tape; the lower filter paper chip slides on the water-based slide frame relative to the upper and middle filter paper chips, so that the different reaction reagents can sequentially flow among the multiple layers of filter paper chips. According to the test card, an early diagnosis device applied in clinical site is provided for patients.

The invention relates to a method for predicting the remaining service life of a universal circuit breaker operating accessory based on statistical data driving. The prediction method comprises the steps that distribution test is carried out on the relative increment of the operating time of the operating accessory from both qualitative and quantitative aspects to verify whether the relative increment conforms to normal distribution; on the basis, the maximum likelihood estimation method is used to estimate parameters in a degradation model; based on the concept of first arrival time, a remaining service life predicting model is established; and an analytical formula of the probability density function of the remaining service life is derived to predict the remaining service life of the circuit breaker operating accessory. According to the invention, the method considers the influence of uncertainties in the operation process of a device on the service life and can acquire the remaining service life distribution required for health management; the method does not use single point estimation; the method is convenient for quantifying the uncertainty of the prediction result of the remaining service life; and the efficiency of the operation and maintenance of a universal circuit breaker can be effectively improved.

The invention discloses a method for evaluating blast vibration damage of a building based on 3D laser scanning. The method uses a 3D laser scanning technology to scan the building, obtains comprehensive point cloud data of the building, and performs processing and modeling to obtain distribution patterns, length and width expansions of surface cracks in all load-bearing parts. A new local crack density is defined, and the local load-bearing structural damage of the building is judged by the local crack density. The statistics of each load-bearing structural damage are conducted to obtain theoverall local crack density so as to evaluate the overall damage of the building. The vibration displacement of a geometric centerline of the building can also be obtained at the same time by using the 3D laser line scanning. This method can also measure the continuity of a target object, and can obtain a real-time damage state of the building structure. The method can quickly give structural damage results in the judgment of building damage in the fields of mine, water conservancy and hydropower, transportation and the like, thereby having a broad application prospect.

The invention discloses an accurate user location identification method based on a neural network. The method comprises the following steps: establishing a position fingerprint database by using collected locations of users and network environments of the users, and then providing accurate locations of the users through the network environments of users having no location. By adoption of the accurate user location identification method, different buildings and different heights in the buildings can be accurately located, the real perception of the users can be accurately reflected, so that we can know the current situation of the network and accurately locate the problem of the present network, an optimization scheme is efficiently output based on matching multiple data sources, and the network problems are discovered prior to customers, thereby reducing the complaints. A lot of manpower and material resources are reduced, meanwhile, a strong support is provided for achieving overall, comprehensive and scientific planning of indoor coverage, the influence of various subjective factors in artificial planning is reduced, and precise planning and optimization of the network can be realized.

The invention relates to a generalized morphological filtering and improved MP algorithm-based low-frequency oscillation mode identification method, and provides a low-frequency oscillation mode identification method of a power system by combining generalized morphological filtering with an improved MP algorithm. De-noising processing is carried out on a low-frequency oscillation signal by using the generalized morphological filtering and then mode identification is carried out by adopting the improved MP algorithm. For the problem of order determination of the MP algorithm in the low-frequency oscillation mode identification process, order determination is carried out by adopting a normalized singular entropy; and the method is relatively small in calculated quantity, high in speed and relatively small in subjective factor influence during order determination. According to the generalized morphological filtering and improved MP algorithm-based low-frequency oscillation mode identification method, accurate identification of a low-frequency oscillation mode under noise interference can be achieved; the problem that the accuracy of low-frequency oscillation mode identification of the power system is seriously affected by noise due to the fact that noise interference is usually introduced in measurement and transmission links of the power system is solved; and the low-frequency oscillation mode identification method has a relatively good application prospect.

The invention relates to a fourth-order mixed mean cumulant and improved TLS-ESPRIT algorithm-based low-frequency oscillation mode identification method. Gaussian color noise is inhibited by using a fourth-order mixed cumulant, a TLS-ESPRIT algorithm is improved, a relative change rate of a singular value is introduced for order determination, and a matrix is constructed by using measured data to carry out low-frequency oscillation mode identification of a power system. According to the fourth-order mixed mean cumulant and improved TLS-ESPRIT algorithm-based low-frequency oscillation mode identification method, detected oscillation frequencies, attenuation factors, oscillation amplitudes and phase information of low-frequency oscillation signals of the power system in different oscillation modes are more accurate.

The embodiment of the invention discloses a method for detecting the quality of experience of video, which comprises the following steps of: acquiring importance information about a video coding data block; packaging the video coding data block into a data transmission packet and generating the importance information about the data transmission packet according to the importance information aboutthe video coding data block; and synchronously transmitting the data transmission packet and the importance information thereof to a video network node, so that the video network node can detect the quality of experience of the video by taking the importance information about the data transmission packet as one of bases for the detection of the quality of experience of the video when the data transmission packet is damaged. Through the method for detecting the quality of experience of the video provided by the embodiment of the invention, the quality of user experience of video services can be relatively more accurately evaluated. The embodiment of the invention further discloses a system and equipment for detecting the quality of experience of the video.

A Bayes and fault tree based reliability evaluation method of a numerical control machine tool belongs to the technical field of reliability evaluation method of the numerical control machine tool. The numerical control machine tool is seen as a system composed of subsystems, and fault data of the subsystem in the same factory batch serves as prior information; a joint probability density functionof failure rates of failure subsystems serves as an onsite data likelihood function, logarithm reverse gamma distribution serves as conjugate prior distribution of the reliability, and a joint priordistribution probability density function of scale and shape parameters of weibull distribution obeyed by fault data is determined; the mean reliability value of the subsystems in a given task time isobtained by using a Bayesian theory; and a fault tree model is established according to connection among the subsystems and influence of the subsystems on a machine tool system, and a fault tree theory is used to calculated the reliability of the numerical control machine tool. The sample amount of prior information can be increased, complex sample compatibility test is avoided, and the compatibility of the prior information is ensured.

The invention provides a skin image information processing-based age testing method. According to the invention, based on the mining of age-related changing texture information of skin surface, a novel, simple and noninvasive age estimation method is provided for medical workers. According to the skin image information processing-based age testing method, a skin image of the middle segment of the ventral side of a forearm is acquired in a noninvasive manner; preprocessing and feature extraction are performed on the acquired skin image, so that relevant parameters of skin grooves and skin ridges are obtained; and the correlation between the parameters and the age is analyzed by using a statistical method, so that the mathematical model of age inference can be obtained. As indicated by statistical results, the skin texture parameters obtained through adopting the method all have high linear correlation with the edge; and when the edge inference model which is obtained through statistical analysis is tested, high accuracy can be obtained. Thus, the effectiveness of the method in age inference can be proved, and cost required by the method is low, and therefore, the method can be prospectively used for objective inference of medical age.

The invention relates to a chestnut drought meteorological index insurance method, which specifically comprises the following steps of: establishing a one-to-one correspondence relationship between meteorological conditions and chestnut yield expected loss values, i.e. Indexing the loss degree to serve as a basis for determining a compensation standard, signing an insurance contract and providingcompensation. The method relates to Chinese chestnut drought disaster risk assessment, meteorological insurance index design, determination of a meteorological insurance claim settlement trigger valueand an insurance rate, development of a meteorological index insurance product and popularization of the meteorological index insurance product. According to the safety method, the precipitation of the Chinese chestnut in the expansion period is determined as the drought index, the insurance party and the insurance party are symmetrical in information, the morality risk and inverse selection in traditional insurance are avoided, investigation and loss assessment by special personnel are not needed, the claim settlement procedure is greatly simplified, and therefore the operation cost of an insurance company is greatly reduced. According to the safety method, the influence of subjective factors in traditional agricultural safety is avoided, farmers are stimulated to actively prevent disasters and reduce damage, the Chinese chestnuts are not damaged due to disasters, and sustainable development of Chinese chestnut planting is guaranteed.

The invention provides a chest radiation-oncology positioning device. The chest radiation-oncology positioning device comprises a fixed frame and a positioning mechanism, a chest-belly positioning groove is formed in the fixed frame, and multiple pressing mechanisms for pressing the positioning mechanism are also arranged at the portions, located on the two sides of the chest-belly positioning groove, on the fixed frame; the positioning mechanism comprises a lower positioning film arranged in the chest-belly positioning groove and connected with the fixed frame, an upper positioning film located on the lower positioning film and connected with the fixed frame and vacuum pads arranged between the upper positioning film and the lower positioning film. According to the chest radiation-oncology positioning device, the lower positioning film is used for positioning the back after a patient lies, the lower positioning film can firmly fit the back of the patient, and the problem that as a gapexists between the back of the patient and the fixed frame, and the patient slightly moves, the treatment effect is influenced is solved; the vacuum pad can also be spliced on the positioning face ofthe upper positioning film or the positioning face of the lower positioning film, the vacuum pads are vacuumized to be formed according to the lying gesture of the patient, the gap between the upperpositioning film or the lower positioning film can be filled with the vacuum pads, and the positioning accuracy is further increased.

The invention relates to a bidding method and system for multi-party comprehensive evaluation and terminal equipment, and the method comprises the steps: determining a first comprehensive score of a tenderbidding unit according to the historical engineering score of a historical cooperation unit for the biddinga tender unit, the integrity score of the bidding unit for the tenderbidding unit, and the professional score of a bidding evaluation expert group for the tenderbidding unit; determining a pre-selected tenderbidding unit from the tenderbidding units according to the first comprehensive score, and sending a score information table to the pre-selected tenderbidding unit; receiving and identifying scoring keywords in the scoring information table, and determining a second comprehensivescore of thea pre-selected tenderbidding unit according to the scoring keywords; and sorting the pre-selected bidding units according to the first comprehensive score and the second comprehensive score, and generating a sorting list. According to the bidding method and system for multi-party comprehensive evaluation, the influence of subjective factors on the bidding process can be reduced to a great extent, and the bidding efficiency and the bidding evaluation quality are improved.

The invention relates to a heuristic rule system based crankshaft wearing detection method for an industrial robot. A heuristic rule system is used for modeling, and a nonlinear mapping relation between variables f1 and f2 (input) of the torque mean value and torque derivative mean value of a crankshaft of the industrial robot and a wearing degree variable Y (output) of the same is described. Vialimited historical data, the initial model of the heuristic rule system is optimized, and influence of subjective factors on the model is reduced. The heuristic rule system method can be used to calculate parameters of the model in a simpler and rapider way, so that the wearing degree of the crankshaft of the industrial robot can be estimated accurately and rapidly by interference of the heuristicrule system under the condition that the torque mean value and torque deviation mean value features are given. The model estimation precision and calculation efficiency are improved, the efficiency is higher, and more advantages are provided.

The invention discloses a coal seam impact risk real-time evaluation method based on a measurement while drilling technology. The method comprises the following steps of synchronously acquiring drilling parameters and drilling cutting quantity data of a drilling machine as sample data while drilling a coal body; establishing a multiple linear regression model, a neural network model and an optimized neural network prediction model; using drilling parameters as input of a prediction model and drilling cuttings as output, comparing and selecting an optimal model as a final prediction model, andautomatically updating the final prediction model to a field construction mobile terminal and an overground computer terminal; when daily impact risk assessment is carried out on a coal seam, analyzing drilling parameters acquired from a pre-assessed coal body automatically by utilizing a final prediction model, so that impact risk assessment is carried out on the coal body. According to the method, the collection work of the drilling cuttings is reduced, the impact risk evaluation procedure of the coal body is greatly simplified, the evaluation time is shortened, and the evaluation cost is reduced.

The invention discloses a method for predicting the average dose of organs at risk in intensity-modulated radiotherapy, comprising the following steps: 1) Obtaining the functional relationship formula between Vox / Vx and Dmx / Dp, wherein, Vx is the volume of organs at risk, and Vox is The intersection volume of the organ at risk and the planned target area, Dp is the prescribed dose, and Dmx is the average dose of the organ at risk; 2) The average dose of the organ at risk is predicted by the functional relationship formula. The specific application of this prediction method includes the following steps: A) determine Dp; B) obtain the average dose Dmx1 of the relevant organs at risk according to the functional relationship formula between Vox / Vx and Dmx / Dp obtained by the prediction method; C) carry out simulation adjustment Strong optimization, get the average dose Dmx2; D) Compare Dmx1 and Dmx2; E) When Dmx2≤Dmx1, meet the standard; Dmx2>Dmx1, continue to optimize the intensity modulation. The invention is easy to collect data, conveniently establishes an organ-at-risk prediction model suitable for corresponding departments' treatment planning systems and equipment, and reduces the influence of subjective factors in the optimization process of radiotherapy plans.

The invention discloses a fuzzy comprehensive evaluation method-based bus passenger satisfaction evaluation method and belongs to the technical field of public traffic control. The method comprises the steps of constructing a bus passenger satisfaction index system; calculating weight vectors of criterion layers and weight vectors of evaluation factor layers through an analytic hierarchy process; dividing satisfaction into L levels and performing value assignment; obtaining the satisfaction of bus passengers for evaluation factors; calculating a satisfaction evaluation matrix of each criterion; calculating the bus passenger satisfaction through a fuzzy comprehensive evaluation method; and establishing a coordinate system by taking the importance of the evaluation factors as horizontal coordinates and the satisfaction evaluation values of the evaluation factors as longitudinal coordinates, and obtaining bus system improvement schemes according to different quadrants in which different evaluation factors are located. According to the method, the more systematic evaluation index system is established; the influence of subjective factors on weights is reduced by applying the analytic hierarchy process; the satisfaction is quantized by applying the fuzzy comprehensive evaluation method; and the improvement measures are proposed through the quadrants in which the evaluation factors are located, so that single and one-sided evaluation of the evaluation factors is avoided.

The invention relates to an electric ovum peeling device and method. The electric ovum peeling device is composed of an ovum peeling tube, an operation handle, a power module, a control module, a display module, a power module, an internal memory, a sucker and / or a voice module and / or a pressure sensor and / or a control box and / or a foot switch controller. All the components can be assembled into an integrated or split type electric ovum peeler. A stepping motor is used for providing blowing and sucking power to peel off granule cells on the periphery of an oocyte. The electric ovum peeling device and method are high in precision, standardization and comfort level and have positive significance in avoiding relevant risks during ovum peeking, achieving standardization of ovum peeling and relieving labor intensity.

The embodiment of the invention provides a business auditing method and device based on machine learning, and relates to the financial technology field. The method comprises the following steps of training a random forest model by adopting the historical business audit data; enabling the random forest model to learn the auditing behavior of an auditor; then inputting the feature vector of a business applicant into each decision tree in the random forest model; and obtaining a classification result outputted by each decision tree in the random forest model, and then determining an audit result of the business request according to the classification result outputted by each decision tree in the random forest model and the weight corresponding to each decision tree in the random forest model. Compared with an expert model, the random forest model is obtained on the basis of the historical business audit data training besides the experiences of the professional audit personnel, so that the dependence on the human experience is small, the influence of the subjective factors is reduced, and the generalization ability and the universality of the audit model are improved.

The invention discloses a method and system for excavation of high-order attributes of test questions. The method comprises the steps that low-order attributes of the test questions and a student's historical answering information are acquired, wherein the low-order attributes include knowledge points or skills of the test questions; the different low-order attributes are combined, and estimated high-order attributes are formed; based on the low-order attributes and the student's historical answering information, the student's ability is determined; according to the student's ability and the student's historical answering information, a weight of the student's ability in each test question based on each estimated high-order attribute is determined; the quantity of the estimated high-order attributes corresponding to the student's ability of which the weights are lower than a set threshold is counted; and if the quantity is larger than a set value, the corresponding estimated high-order attributes are taken as the high-order attributes of the test questions. With the method and system disclosed by the invention, the high-order attributes of the test questions are determined efficiently and accurately.

The invention discloses a poor user identification method based on daily differential electricity consumption, and relates to a poor user identification method. At present, a user group type is subjectively defined often directly according to a clustering result, so that a relatively large error exists in user group classification. The method comprises the steps: firstly, carrying out poor user statistics on an investigated user group in a questionnaire mode, then dividing the investigated user group into a training sample group and a test sample group, and meanwhile, carrying out daily electricity consumption difference calculation on users of the training sample group to obtain the daily difference electricity consumption data; and respectively extracting electricity utilization characteristic data of poor users and non-poor users from the daily differential electricity consumption data by utilizing a statistical method. On the basis, the test sample group is verified, and a poor user identification method is perfected. According to the technical scheme, the influence of subjective factors is reduced, and the user group classification accuracy is effectively improved; the recognition efficiency is high, the coverage range is wide, and a large amount of manpower and material resources are saved.

The invention discloses an ultrasound breast tumor grading method based on multi-feature extraction and Linear SVM. The method comprises steps that to-be-diagnosed breast ultrasound RF data is inputted; the RF data is converted into an ultrasound B-mode image to obtain the initial suspicious lump area location; the initial suspicious lump area is segmented to obtain a suspicious lump to determinethe boundary of the suspicious lump; a feature value of the segmented suspicious lump is calculated; the calculated feature value is inputted into a classifier, and the suspicious lump area is analyzed. The method is advantaged in that based on the BI-RADS grading standard, the actually acquired breast ultrasound RF data is hierarchically detected, breast tumor ultrasound 3, 4 and 5-level classification is achieved through utilizing the multi-feature extraction method and the linear support vector machine classifier, improvement of accuracy of the doctor's diagnosis is facilitated, the algorithm time and complexity can meet clinical requirements, and the method has a certain practical value.