8 results about "Factor selection" patented technology

The invention provides a process parameter optimization method and device. Irrelevant noise factors are removed under data driving through a dynamic clustering algorithm, core factors are reserved, and blindness of factor selection is avoided; the improved optimization algorithm uses a dynamic penalty function mechanism, and a search path of the algorithm is forced to be always kept in a high-quality safety region, so that the occurrence of a process parameter combination causing defects is effectively prevented, and the application effect of the process parameter combination is improved; the experiment matrix is generated by combining the split area experiment structure designed by the parameter adjustment cost of each core factor, so that the experiment time consumption can be reduced, and the experiment efficiency is improved; dynamic adaptation to experimental data fluctuation is achieved through dynamic feasible region constraint of core factors, it is guaranteed that experimental points are in a safe interval, and the application effect of technological parameter combination is guaranteed.

According to the federated learning optimization method based on the self-adaptive loss threshold, the threshold is dynamically calculated and adjusted according to the error change trend in the training process, so that the client selection strategy is optimized. Secondly, the invention provides a two-factor selection strategy which can select different selection strategies according to different training stages. And finally, a historical information interaction mechanism is provided, so that the client selection is not only based on the current performance index, but also can be flexibly adjusted according to the change trend of the historical performance. According to the mechanism, the client selection process can adapt to the requirements of different training stages, and the training stability and convergence efficiency can be remarkably improved. Experimental results show that compared with a Pow-d method, the global model accuracy rates of the AHIBCS on the three public data sets of COVID-19, FMNIST and CIFAR-10 are improved by 2.1%, 2.4% and 2.7% respectively.

The application relates to a prompt word optimization method and device of a text evaluator and a storage medium, and belongs to the technical field of artificial intelligence. The application first initializes a selection strategy cluster, which comprises a plurality of design factor selection strategies of an evaluation prompt word. In each iteration, each selection strategy in the selection strategy cluster is disturbed to generate a new selection strategy. The evaluation prompt word is determined based on the new selection strategy, and an evaluation result is generated on a verification set with artificial evaluation by using a large language model. The correlation coefficient of the evaluation result and the artificial evaluation is calculated, and based on the correlation coefficient, the selection strategy cluster is updated from the current selection strategy cluster and the new selection strategy. The application adopts an iterative search method guided by a heuristic function to optimize the selection strategy, and the selection strategies of a plurality of design factors in the prompt word are optimized, the search range of the prompt word is expanded, and the evaluation performance of the text evaluator is improved.

This disclosure relates to the field of information technology, and in particular to a method and system for implementing an arrangable project development scaffold. The method includes: a factor management module that creates, configures, and publishes multiple scaffolding factors for a project to be developed, and provides a factor management library for storing these multiple scaffolding factors; a template creation module that provides a scaffolding template library storing various preset scaffolding templates, and, in response to a user selecting at least one preset scaffolding template from the library for the project to be developed, and performing factor selection and factor arrangement operations on multiple scaffolding factors within the at least one preset scaffolding template, generating a target scaffolding template; and a scaffolding generation module that, in response to the user's selection of the target scaffolding template, generates a target scaffolding based on the target scaffolding template, so as to utilize the target scaffolding to perform development tasks. This disclosure automates and automates the entire development process through software scaffolding, thereby improving development efficiency and reducing repetitive work.

Provided is a method for narrowing down factors that are used in state evaluation. This factor selection device comprises: a data acquisition unit that acquires candidates for factors that are used in assessing an object; an evaluation unit that evaluates the magnitude of impact upon assessment with regard to each factor; and a factor selection unit that, on the basis of the result of evaluation from the evaluation unit, selects a factor having a high impact upon the assessment. The evaluation unit selects one unevaluated factor from among the candidates, performs an assessment excluding the factor, and evaluates the accuracy of the assessment. If there is a change in the accuracy of assessment, an evaluation process to evaluate the relevant factor as an impacting factor is repeatedly executed for each of the candidate factors, and the factor selection unit selects a remaining factor.

The application provides a GNSS / SINS combined navigation position variance elasticity adjustment optimization method, system, device and medium based on deep learning, belonging to the navigation field, comprising a multi-dimensional precision characteristic factor selection step, a convolutional neural network training step, a variance elasticity adjustment and filtering solution optimization step; the multi-dimensional precision characteristic factor selection step determines the parameter index associated with the position error through a nonlinear correlation calculation method, the convolutional neural network training step performs offline training of the position precision prediction model based on the selection result of the parameter index, and the variance elasticity adjustment and optimization step uses the offline trained model to predict the position precision online and elastically adjust the position variance, so that the true error of the position state in the filtering solution and its variance are more matched, thereby making the prior information and the current information obtain a more reasonable weight ratio, and improving the current combined navigation positioning and pose performance in a complex electromagnetic environment.

The present application relates to multi-factor selection of multi-scheme, and provides a multi-factor structural weight analysis method for multi-scheme selection, which is used for determining the influence degree of multi-factor on the selection process of multi-scheme. Since the scheme has not been implemented, the scheme implementation results under the influence of multi-factor are lacking. The method represents the integrity of factor influence through the relationship between the factor value domain and the whole domain of the factor value in each scheme. The distinctness of factor influence is represented through the value domain relationship of the same factor in different schemes. The integrity and distinctness are described by interval numbers, and the factor structural weight is formed by defining the integrity degree and distinctness degree. The method includes the construction of an analysis system, a decision matrix, a normalization matrix, the determination of the factor integrity degree and distinctness degree, the factor influence degree and weight. The method is applied to the determination of the structural weight of each factor in the selection of mining methods. The method can be applied to the determination of the influence of factors on the selection of schemes under the condition of lacking scheme results.

The invention belongs to the technical field of expert systems, and particularly relates to a centralized control expert system model construction method, which comprises the following steps of: screening candidate factors obviously related to a target variable from time sequence correlation characteristics, and combining contribution degree analysis and a weighted priority distribution mechanism to realize dynamic sorting of factor importance; on the basis, multiple groups of candidate structures are generated and cross validation is performed, so that the stability and generalization performance of different factor combinations are effectively evaluated; and finally, the system can continuously optimize the model structure according to evaluation feedback by iteratively correcting the factor weight and the combination rule. According to the method, the factor identification capability and the model adaptability of an expert system in a complex and dynamic data environment are remarkably improved, the problems of inaccurate regulation and control, response delay and the like caused by factor selection deviation and weight setting rigidity of a traditional system are solved, and therefore the intelligent decision-making level and the operation stability of the whole system are improved.