80results about How to "Improve fairness" patented technology

The invention belongs to the technical field of cosmetics, and particularly relates to a skin-color-uniformizing whitening composition derived from plant extracts and application of the skin-color-uniformizing whitening composition. The whitening composition provided by the invention is prepared from the following components in parts by weight: 10 to 15 parts of saussurea involucrata extract, 1 to 8 parts of dendrobium officinale extract, 1 to 5 parts of wood apple extract and 0.5 to 5 parts of banana fruit extract. The four plant extracts are synergistically compounded, transdermal absorption of active ingredients can be promoted, the tyrosinase inhibition rate is effectively increased, and the skin glossiness and whiteness are improved, so that the effects of whitening, brightening and uniformizing the skin color are achieved, the whitening composition is mild and non-irritant, the preparation method is simple and convenient to operate, and the whitening composition has large-scale production capacity and industrial feasibility. When the whitening composition is applied to skin care products such as face cream, the whitening and brightening effects of the face cream can be effectively improved, and the whitening composition has wide market application value.

The invention discloses an urban waterlogging resident emergency evacuation path guiding method for a smart community, and relates to the technical field of smart cities and emergency management. The method comprises the following steps: firstly, constructing a community digital twinborn base fusing geographic information, building information, Internet of Things sensing data and a hydrodynamic model; dynamic inversion and prediction of the waterlogging situation are realized by assimilating a multi-source observation data driving model; calculating the space-time dynamic risk of the evacuation road network based on the prediction result, and generating a risk map; calculating a space-time evacuation scheme including optimal departure time and a specific path sequence for residents or families by combining resident personalized constraints and adopting a double-layer optimization model combining macroscopic flow distribution and microscopic path planning; finally, instructions are issued through multiple channels such as an intelligent terminal and broadcast, and dynamic adjustment is performed based on real-time monitoring, so that accurate perception, prospective prediction, personalized planning and collaborative guidance of waterlogging evacuation are realized, and the overall efficiency of community emergency evacuation is effectively improved.

The invention relates to the technical field of power system carbon emission analysis, and discloses a carbon flow closed-loop traceability and allocation method, system, equipment and medium considering network loss, which comprises the following steps: analyzing network loss components contributed by each load, and establishing quantitative correlation between load power consumption behaviors and network loss power; and then, constructing an equivalent lossless network, reasonably allocating the carbon emission corresponding to the network loss to a load side, and introducing a renewable energy carbon reduction load mechanism to quantify the carbon emission reduction benefit of the renewable energy. The method aims to improve the accuracy and rationality of carbon flow allocation in the whole link, and provides reliable support for low-carbon scheduling and carbon market mechanism design of a power system.

The application provides a low-cost two-wire serial communication method for multi-device competition, which adopts two signal lines to realize bus arbitration and data transmission among multi-devices, wherein any signal line is used as a chip selection line, another signal line is used as a bidirectional data line, chip selection pins of all devices are configured as open-drain output, the chip selection line is connected to a positive electrode of a power supply through an external pull-up resistor, and a high level in a normal state represents a bus idle; when multiple devices share the same bus, distributed arbitration is realized through the following steps: a device continuously monitors a state of the chip selection line before sending data; if the state of the chip selection line is a high level, the device pulls down the chip selection line to attempt to obtain control right of the bus; if the state of the chip selection line is a low level, the device waits until the chip selection line is released; the device obtaining the sending permission sends data with a self-defined frame structure through the bidirectional data line; after the sending is completed, the device obtaining the sending permission releases the chip selection line.

The invention provides a post-processing federated long-tail classifier calibration method and device based on random noise driving, and the method comprises the steps: carrying out the multi-round federated learning training between a central server and a plurality of clients, and carrying out the aggregation to obtain a pre-training global model comprising a feature extractor and a classifier; at the central server side, freezing the feature extractor, and constructing a virtual balance data set based on random noise sampling and uniform label sampling; optimizing parameters of the classifier by using the virtual balance data set and a calibration loss function to obtain a calibrated classifier; and for a to-be-classified test sample, predicting the to-be-classified test sample by using the feature extractor of the central server side and the calibrated classifier. The method is applied to an image classification task in federated learning, classifier deviation caused by long-tail distribution can be effectively calibrated under the conditions that data privacy is completely protected and no auxiliary information is needed, and the recognition performance and the overall prediction calibration degree of the model on the tail category are remarkably improved.

This invention discloses a dynamic adaptation method for organic agricultural product consumption scenarios based on multimodal perception. The method is executed by an edge server in the exhibition hall and includes: receiving a display collaboration request sent by a mobile terminal; determining a target video based on the source identifier of the streaming media content and product information; determining whether multiple public display terminals are available; and if a first public display terminal is determined to be available, sending the target video to the first public display terminal for playback and sending the terminal identifier of the first public display terminal to the mobile terminal to notify the user of the first public display terminal. This invention can improve the collaborative display efficiency between mobile streaming media content and public display resources in organic agricultural product exhibition halls and enhance the dynamic adaptation effect in agricultural product consumption scenarios.

The present application relates to the technical field of power system operation optimization, and particularly relates to a peak shaving auxiliary service cooperation game cost allocation method and system considering environmental externality, which comprises obtaining system time series data and unit operation emission data, determining the net load fluctuation of a peak shaving demand subject according to the system time series data, and generating an environmental cost item representing the environmental externality of deep peak shaving of a thermal power unit according to the unit operation emission data.The present application determines the net load fluctuation of a peak shaving demand subject and the environmental cost item representing the environmental externality of deep peak shaving of a thermal power unit by obtaining system time series data and unit operation emission data, and introduces the environmental cost item into the dispatch optimization target of joint clearing, so that the joint clearing model can simultaneously consider the system operation cost and the environmental cost generated by deep peak shaving when optimizing the dispatch strategy.

The present application belongs to the technical field of Internet of Vehicles and communication security, and discloses a federated learning training method based on vehicle state perception and applicable to dynamic Internet of Vehicles. The method collects vehicle running state information of vehicle clients, preliminarily screens candidate vehicle clients, calculates comprehensive scores of the screened vehicle clients, adopts a client selection strategy combining selection of the top-ranked clients with random supplementation to determine a set of vehicle clients participating in the current round of federated learning, configures local training parameters for the selected vehicle clients according to vehicle speed and computing capacity, performs local training on local data sets to obtain model updates, performs weighted aggregation on the model updates of the effective participating vehicle clients at the server end, introduces a momentum mechanism to smooth the model update process, and obtains new global model parameters. This method effectively improves the stability, robustness and training efficiency of federated learning in a high-dynamic Internet of Vehicles environment.

The application discloses an employee performance automatic examination and evaluation management system, and relates to the technical field of intelligent performance evaluation.The system comprises a structured data collection module, a behavior data modeling module, an initial performance modeling module, a dynamic weight optimization module, a performance report generation module and a blockchain storage module.The structured data collection module is used to connect the human resource information system, the enterprise resource planning system, the customer relationship management system and the project management platform in the enterprise through a standardized interface, and to collect the structured performance-related data generated by the employees in the above systems in real time based on a business event triggering mechanism.The multi-source structured and unstructured data are fused to construct a collaborative relationship modeling mechanism based on a graph neural network and a dynamic weight optimization mechanism based on reinforcement learning, so that a comprehensive, objective and personalized evaluation of the employee performance is realized.Meanwhile, the blockchain storage technology is introduced to ensure that the whole examination process is traceable and tamper-proof.

The application discloses a method and device for inhibiting strategic bidding of power suppliers under a partition settlement mechanism, and the method is established as a three-layer collaborative model: an upper-layer model takes a candidate partition scheme as a decision object, enumerates and compares multiple candidate partition schemes, and determines an optimal partition scheme; a middle-layer model takes the candidate partition scheme as a basis, takes maximization of power supplier revenue as a target, determines a power supplier bidding coefficient through Nash equilibrium iteration, forms a bidding strategy; and a lower-layer model takes the partition scheme and the bidding strategy as a basis, takes minimization of total power generation cost as a target, performs market clearing, obtains unit output, node marginal price and partition settlement price, and feeds back the results to the upper-layer model and the middle-layer model, so as to drive scheme optimization and strategy iteration. The method aggregates node price signals through the partition settlement price, weakens the ability of power suppliers to manipulate local prices, inhibits strategic bidding, alleviates pseudo congestion, reduces system operation cost, and improves market efficiency.

This invention discloses a method and system for optimizing community bus service zones and sub-zone routes based on connection resistance, comprising the following steps: S1, constructing a main model for community shuttle bus service zones based on finite center-subordination rules to obtain the service sub-zone allocation scheme for community shuttle bus stops; S2, for each service sub-zone, constructing a sub-model for optimizing community shuttle bus service routes based on bus service impedance index to obtain shuttle bus service route schemes for each sub-zone connecting to public transportation hubs, determining whether the minimum service standard for shuttle buses is met by checking the connection resistance; for all sub-zones that fail the connection resistance test, adding the corresponding sub-zone partitioning constraint set to the main problem, and returning to S1 to resolve the main model until all shuttle bus route schemes for each sub-zone pass the test. This invention, through connection resistance testing, makes the designed community shuttle bus connection routes more competitive in service compared to shared bicycle travel.

The invention discloses a data processing method and device based on an alliance chain, equipment and a readable storage medium. The method comprises the steps of obtaining a target consensus type corresponding to a target alliance party to which a first consensus node belongs; performing consensus voting processing on at least one consensus node contained in the target alliance party according to a consensus voting rule indicated by the target consensus type; the consensus voting processing is used for electing voting representative nodes of the target alliance party from at least one consensus node contained in the target alliance party; the voting representative node of the target alliance party is used for representing the target alliance party to carry out voting consensus processing on the proposal block in the alliance chain. By adopting the method and the device, the decentralization and the fairness of the alliance chain can be improved.

The invention discloses a new energy transaction settlement deviation compensation method, device and equipment and a computer program, and the method comprises the steps: determining transaction settlement decision variables of a new energy generator, an electricity selling company and a power user, and determining a settlement deviation variable; constructing a power generator objective function, a power selling company objective function and a user objective function, and determining compensation coefficient sensitivity indexes of the new energy power generator, the power selling company and the power user; constructing a deviation compensation optimization model based on game equilibrium, and constructing a compensation coefficient dynamic adjustment model; and solving the deviation compensation optimization model through a particle swarm algorithm fusing a chaos particle swarm and Nash equilibrium guidance to obtain an optimal solution of a transaction settlement decision variable, determining an optimal compensation coefficient according to the optimal solution and a compensation coefficient dynamic adjustment model, and further performing new energy transaction settlement deviation compensation. The accuracy, fairness and real-time performance of new energy transaction settlement deviation compensation are improved, and the method can be applied to the technical field of electricity markets.

This invention relates to the field of behavioral profile generation technology, specifically disclosing a method for generating user behavior profile content based on machine learning, including the following steps: Step S1: Collect historical behavior logs of multiple users, count the total interaction frequency of a single user, distinguish active and inactive users based on thresholds, and construct frequency vectors for each user's behavior items; Step S2: For a single user, calculate the angle between its frequency vector and the vectors of all active users, select the K active users with the smallest angle, aggregate their vectors and concatenate them with the original vector of the target user to generate an enhanced behavioral feature vector; Step S3: Utilize two sets of linear mapping matrices to operate on the enhanced features respectively to obtain frequency and interest representation vectors, calculate the mean cosine similarity between the two, iteratively optimize the classification based on this value, determine the final interest representation vector, and output the user behavior profile text accordingly.

The present application relates to the technical field of power market data analysis, solves the problem that the prior art lacks a verification mechanism for the credibility of generated reports and the traceability of evidence, and is also prone to generating false content that contradicts the actual physical operating state, thereby affecting the credibility of the report, and particularly relates to a power regulation report credibility verification method based on data traceability, which collects and preprocesses multi-modal operating data of the power market to obtain baseline operating data, then performs multi-dimensional feature extraction thereon to generate a multi-modal basic feature vector, assigns a traceability identifier to each valid data entity in the multi-modal basic feature vector, and synchronizes the data entity and its corresponding data feature abstract to a distributed tamper-proof network to construct a traceability credential library. The present application improves the evidence traceability of the analyzed text, reduces the risk of tampering with the operating data during circulation, and improves the reliability of the analysis and periodic regulatory evaluation results of the market operation.

The invention discloses an intelligent talent recruitment screening method based on multi-source data fusion, and relates to the technical field of big data management, and the method comprises the steps: collecting candidate multi-source heterogeneous data, carrying out the feature extraction, and generating a multi-modal initial feature vector; fairness related attribute data and historical recruitment result data of candidates are collected and combined with the skill confidence feature set, and a recruitment decision causal graph model is constructed; and based on a recruitment decision causal graph model, performing causal effect estimation and anti-fact reasoning by taking the multi-source verification confidence as a weight, performing fair correction calculation on candidate ability evaluation represented by the skill confidence feature set, and outputting a fair correction talent sorting list. According to the method, the multi-source evidence cluster is constructed in the shared semantic space, and the multi-source verification confidence is calculated, so that the unified management and credibility quantification of the candidate multi-source heterogeneous data in the skill dimension are realized, and the controllability and traceability of a skill assessment result in a big data management scene are improved.

The invention discloses a cross-data center RDMA network congestion control method and device based on in-band telemetering, and belongs to the technical field of network transmission control. The method comprises the following steps: sending a first cross-data center flow to an upstream DCI switch at an initial rate; according to a Switch-INT message returned by an upstream DCI switch, a first fair rate is calculated, a second cross-data center flow is sent to the upstream DCI switch at the first fair rate, an ACK message of a downstream data center responding to the first cross-data center flow is obtained, and the ACK message comprises a suggested rate; and generating a final rate in combination with the first fair rate and the suggested rate, and sending a third cross-data center flow to an upstream DCI switch at the final rate. According to the method, the network sensing capability of the cross-data center flow can be improved, the fairness of the flow in the data center and the cross-data center flow is improved, and the transmission delay of the datagram is optimized.

This invention discloses a trusted auditing method for the operation and maintenance of an points mall based on a blockchain distributed ledger, comprising the following steps: Step 1: Monitoring the operation event flow in the points mall, collecting points operation behavior data, and parsing the event types; Step 2: Using an improved SHA-256 hash algorithm to form a points operation behavior evidence record; Step 3: Constructing an operation behavior graph based on the points operation behavior evidence record; Step 4: Performing numerical processing to generate behavior representation vectors; Step 5: Performing feature analysis, extracting explicit risk factors, and forming a risk score; Step 6: Generating interpretable audit records for operation paths exceeding a preset risk threshold, and performing digest compression; Step 7: Writing the structured audit evidence data into the blockchain ledger and performing manual review. This invention integrates on-chain evidence storage and behavior scoring mechanisms to achieve trusted auditing and risk control supervision throughout the entire points operation process.

This invention relates to the field of data evaluation technology, and discloses a multi-source mapping data fusion evaluation method and system for agricultural insurance. The method includes: acquiring images of a target planting area before and after the snowmelt period; extracting the reflectance difference of the snow spectrum of each plot unit in the images under shaded and unshaded conditions to obtain the change in snow cover rate of the plot unit per unit time; identifying accelerated ablation blocks in the target planting area based on the change in snow cover rate; and constructing a snowmelt runoff catchment tree based on the topological connection between the water runoff direction of the slope where the accelerated ablation block is located and the downstream plots; based on the snowmelt runoff catchment tree, this invention can solve the problems of existing technologies having ambiguity in determining the range of runoff scour caused by snowmelt, lacking a nighttime temperature correction mechanism for runoff freezing, and having large deviations between simulated snowmelt flow and actual conditions.

The application discloses a GPU-based multi-client real-time spectrum monitoring system and a method thereof, and belongs to the technical field of spectrum analysis.The method comprises the following steps: acquiring the spectrum state, behavior information, host state and service quality requirement of a client; calculating the interference intensity coefficient, behavior state coefficient, host state coefficient and service quality coefficient of the client through a corresponding model; dynamically allocating the computing power of each client by using a power distribution model based on the above coefficients; and finally outputting the target refresh rate of each client by a refresh rate optimization model in combination with the allocated power and service quality coefficient.The application solves the problem that the resource allocation in the prior art is rigid and cannot adapt to the differentiated requirements of multi-clients by multi-dimensional state sensing and modeling, realizes intelligent, dynamic and efficient allocation of system computing resources, guarantees the service quality of multi-services, and improves the overall energy efficiency and stability of the system.

The invention relates to the technical field of artificial intelligence, and provides a composition scoring method and device, a medium, electronic equipment and a program product, and the composition scoring method comprises the steps: obtaining a to-be-scored composition, and determining the semantic information of the to-be-scored composition; according to the semantic information of the to-be-scored composition, determining a logic dimension score, a vocabulary dimension score, a grammar dimension score, a content dimension score and an expression dimension score of the to-be-scored composition; and performing weighted summation on the logic dimension score, the vocabulary dimension score, the grammar dimension score, the content dimension score and the expression dimension score to obtain a composition score of the to-be-scored composition. The composition scoring is automatically, comprehensively and undifferentially carried out, the cost is reduced, the efficiency is improved, and the impartiality and consistency of composition scoring are improved.

This application provides an operation processing method, apparatus, device, and readable storage medium. In the technical solution provided by this application, a dual-level processing mechanism of message level and point level is constructed. At the message level, permission conflict detection based on user identity is introduced to effectively prevent non-creator users from accidentally deleting or tampering with other people's content, thus achieving fine-grained operation isolation. At the point level, conflict detection is pushed down to the smallest data unit coordinate point. By comparing the point position sets of the current operation and the historical operation, the spatial overlapping area is accurately identified, and the operation type corresponding to the current operation message is combined to dynamically determine whether execution is allowed, ensuring that the operation result is certain and predictable. This achieves hierarchical deterministic operation conflict processing, improving the smoothness and user experience of the online collaborative whiteboard system.

The invention relates to the technical field of attendance management, and discloses a technical support data intelligent analysis system, which comprehensively collects feature facility operation and key point road condition information through a region division module and a working region unit key information collection module, and provides rich data support for attendance analysis. The attendance checking terminal corresponds to the working area unit, so that the attendance checking state and time of employees can be obtained in time, and the accuracy of attendance checking data is guaranteed; specific working areas of employees are defined through the employee information storage module, so that targeted management is facilitated; the arrival time of a regular bus and a staff list are accurately recorded through the regular bus information acquisition module, and a key basis is provided for card punching correction analysis; through mutual cooperation of the judgment module and the analysis module, whether card punching correction is needed or not can be intelligently judged, correction is automatically completed, attendance checking errors caused by regular bus factors are avoided, it can be ensured that cards are supplemented only when employees arrive at a working area at reasonable time, effective performance management of enterprises is facilitated, and the overall management fairness is improved.

The invention discloses an image generation model-oriented watermark generation and detection method and system, and the method comprises the steps: a model provider pre-embeds an invisible watermark in a diffusion model, and provides a high-quality non-perception image generation service for a legal user; when a malicious user generates and propagates illegal or harmful images by using an API, a detection request can be provided to a model party through a supervision party, and a model provider performs watermark detection on suspicious images; and the model provider extracts watermark information from the suspicious image, judges whether the watermark information is generated by the model or not, and submits a detection result to the supervisor to judge whether the watermark information is legal or not. According to the method, a watermarking mechanism which is only used for'detection 'instead of'traceability' is designed, the speech freedom and privacy of legal users are protected while the identifiability requirement of supervision on AI generation content is met, direct tracking of user identities is avoided, and therefore effective balance between content safety guarantee and user right maintenance is achieved, and user experience is improved. And the fairness and acceptability of AI content governance are obviously improved.

The application discloses a fair personalized recommendation method and device based on mutual information decoupling and a storage medium, and steps of the method comprise the following steps: 1. constructing original data, including a user-product rating matrix and a user sensitive attribute matrix; 2. constructing a sensitive embedding network to learn sensitive embedding of a user and a product, including a biased one-hot encoding layer, a sensitive information encoder and a sensitive attribute prediction layer; 3. constructing a hybrid embedding network to learn hybrid embedding of the user and the product, including a hybrid one-hot encoding layer, a hybrid information encoder and a preference prediction layer; and 4. constructing a non-sensitive embedding network to learn non-sensitive embedding of the user and the product, including an unbiased one-hot encoding layer, a mutual information lower bound optimization layer and a mutual information upper bound optimization layer. Through the double mutual information fairness constraints on the embedding vectors, the application improves the fairness of the recommendation system while ensuring the recommendation accuracy.

The application discloses a mahjong machine with a control lighting lamp, and belongs to the technical field of mahjong machines, and has the technical scheme as follows: the mahjong machine comprises a mahjong machine main body, irradiation lamp strips are rotationally arranged on four sides of the mahjong machine main body, placing areas for placing mahjong hands are arranged on the four sides of the mahjong machine main body, the light beams of the irradiation lamp strips are obliquely shot from bottom to top to the mahjong hands in the corresponding placing areas, a controller is arranged on the mahjong machine main body, human body sensing modules are arranged on the four sides of the mahjong machine main body, the human body sensing modules are electrically connected with the controller, the controller is electrically connected with the irradiation lamp strips, when the human body sensing modules sense a human body, the controller controls the corresponding irradiation lamp strip to be turned on, and the four sides of the mahjong machine main body are provided with adjusting knobs for adjusting the brightness of the corresponding irradiation lamp strips, the mahjong machine forms effective light zones on the mahjong hands through the irradiation lamp strips, the recognition of the mahjong hands by users is enhanced, and the use experience is improved.

The application discloses a charging and discharging strategy scheduling method and system of an electric vehicle, relates to the technical field of vehicle network interaction and intelligent power grid scheduling, and comprises the following steps: acquiring current micro power grid area real-time scheduling related data; aggregating the first two layers of parameters of a Critic network in an initial scheduling model of all vehicles in the area to obtain global parameters; updating corresponding parts of local models of the vehicles based on the global parameters to form a collaborative scheduling model; after the models generate charging and discharging strategies based on real-time data, the charging and discharging strategies of each vehicle are output after competitive optimization according to the willingness of users to participate; and the application realizes collaborative optimization while fully protecting the privacy of users by aggregating model parameters instead of sharing original data; the willingness driving mechanism is introduced to improve the acceptability and execution rate of the scheduling scheme; and the adaptive scheduling capability of the system to dynamic environments such as electricity price fluctuation and uncertain user behavior is enhanced by combining federated learning and reinforcement learning.

The application relates to a blockchain consensus method and device, a computer device, a storage medium and a computer program product. The method comprises the following steps: determining an initial node set in a blockchain; broadcasting, by each node, a first value in a respective value chain in a reverse order, and integrating the first values broadcast by each node to obtain a target random number if each node has broadcast the corresponding value within a first preset time length; determining a first block node from all nodes in the initial node set according to the difference between the first value broadcast by each node and the target random number; obtaining a new target random number according to the first block node, determining a next block node according to the new target random number, and stopping until the total number of blocks reaches a preset number. The method can improve the fairness of selecting a block node in a blockchain system.

The application relates to a visual language model-based incremental learning method and system, and relates to the technical field of computer vision. The method comprises the following steps: fine-tuning a visual language pre-training model according to training sample data and historical sample data of a current task to obtain a target visual language model; extracting discriminative image features of all sample data according to the target visual language model; determining class feature means and a shared covariance matrix of a plurality of known categories involved in the current all sample data respectively according to the discriminative image features; dynamically calibrating the experience frequency of the plurality of known categories according to an inverse frequency adjustment mechanism to obtain calibrated prior probability; determining a debiased classification parameter according to the calibrated prior probability, the class feature means and the shared covariance matrix to construct a target debiased classifier; and performing weighted summation on the output result of the target visual language model corresponding to a to-be-classified image and the output result of the debiased classifier to obtain a target classification result.

The application relates to the technical field of satellite communication, and discloses a transmission resource scheduling method, equipment and a program product, wherein the method comprises the following steps: obtaining one or more target transmission requests, the target transmission requests coming from different target equipment; determining a scheduling control coefficient of each target transmission request, and constructing a transmission request sequence based on the scheduling control coefficient, wherein the scheduling control coefficient comprises a load saturation coefficient and a forwarding priority coefficient; obtaining a channel quality parameter corresponding to each target transmission channel, determining an available waveform set of the target transmission channel according to the channel quality parameter, and determining a target transmission waveform of the target transmission request in the available waveform set; for any target transmission request in the transmission request sequence, allocating a time slot resource to the target transmission request according to the target transmission waveform, and distributing the allocation result to the corresponding target equipment. The technical scheme provided by the application can improve resource efficiency, fairness of resource allocation and reliability of transmission.