9 results about "Indicator vector" patented technology

The application provides an engineering education system and method fusing explainable multi-objective intelligent optimization, and relates to the technical fields of intelligent education and artificial intelligence optimization. The method comprises the following steps: obtaining an initial engineering design scheme of a user, and obtaining a standardized multi-dimensional index vector through physical simulation analysis and logical analysis; configuring weights with structured reasons for each index, identifying critical weight values leading to decision reversal through sensitivity analysis, and then determining optimization objectives and constraint conditions; performing multi-objective optimization search using an explainable multi-objective optimizer, generating a Pareto optimal solution set and outputting explainable data; simultaneously generating a behavior interaction sequence of the user; generating a capability evaluation report of the user according to the behavior interaction sequence; and adjusting a preset teaching logic sequence according to the capability evaluation report to form a closed-loop teaching process. The application is suitable for cultivating and evaluating the decision-making ability of students in balancing technical performance and environmental sustainability in engineering design.

This invention discloses a dynamic interactive response method and system for e-commerce applications based on sensor data, comprising the following steps: collecting and preprocessing data from the e-commerce application's operation to generate a standardized interactive input data set; extracting continuous state features and discrete event features to construct a dual-path coupling control sequence; performing stage correlation calculations to generate interactive stage indicator vectors; inputting the dual-path coupling control sequence and interactive stage indicator vectors into an improved NCDE model to generate a dynamic interactive hidden state sequence; performing response sensitivity modulation to construct a dynamic interactive response strategy set; dynamically adjusting the interface based on the dynamic interactive response strategy set to generate dynamic interactive response results; collecting subsequent feedback and writing back updates to optimize the dynamic interactive hidden state sequence and the dynamic interactive response strategy set. This invention improves the real-time performance, accuracy, and adaptive optimization capabilities of e-commerce interactive responses.

This invention belongs to the field of sensors and signal processing, specifically relating to a wavelet kernel-scale sensitivity-guided denoising method for abrasive particle induced voltage signals. The method includes: acquiring abrasive particle induced voltage signals and performing harmonic cancellation to obtain a preprocessed signal; constructing a wavelet kernel function and calculating a kernel-scale guided spectrum using it and the preprocessed signal; constructing a sparse joint denoising model based on the kernel-scale guided spectrum; processing the sparse joint denoising model to obtain a convex optimization objective function; solving the convex optimization objective function using an adaptive step-size gradient descent method and an adaptive iterative shrinking threshold method to obtain a weight vector characterizing the distribution of abrasive particle characteristic signals; binarizing the weight vector characterizing the distribution of abrasive particle characteristic signals to obtain a feature indicator vector; performing a Hadamard product between the feature indicator vector and the preprocessed signal, followed by low-pass filtering to obtain a denoised signal. This invention can adaptively and non-destructively enhance and denoise abrasive particle characteristic signals under strong interference environments.

The application relates to a key region identification method based on a four-dimensional characterization index. The method comprises the following steps: based on a sample set, calculating an information surface entropy, a maximum gradient consistency, a local-global similarity divergence and a local-global extreme point ratio to form a four-dimensional characterization index vector of a key region; fusing the four-dimensional characterization index vector based on a Pareto dominance relationship, dividing candidate key regions into different front levels through non-dominated sorting, calculating the crowding distance of each front region, and screening the key region in the test sample space in combination with the front level and the crowding distance. The method can improve the utilization efficiency of a new type of aero-engine digital twin test resources.

This invention provides a method, application, and electronic device for constructing an author name disambiguation model. First, a name block and an initial training set consisting of entity pair training samples are constructed. An author name disambiguation baseline model is trained based on the initial training set. The baseline model is then used to calculate the global feature importance vector for each entity pair training sample through feature attribution calculation. The score for each entity pair training sample is calculated based on the global feature importance vector and a non-empty indicator vector, and a training set is selected. The baseline model is then retrained using the selected training set to obtain the author name disambiguation model. The objective loss function of the author name disambiguation model includes classification loss and attribution prior regularization, improving model interpretability, reducing information loss due to metadata sparsity, suppressing model attention drift, and being compatible with various model structures. It has low engineering modification costs and can be widely applied to name disambiguation scenarios in various academic data processing systems.

The application discloses a kind of observation record generation methods, systems, equipment and medium of infant behavior, which comprises: first, the real-time behavior video stream of infant is acquired and preprocessed, and the multi-modal feature vector containing action, emotion, environment / object and identity is extracted by lightweight neural network;Then, the index of "3-6 years old children learning and development guide" is converted into a standard index vector group, and the semantic correlation score of features and indexes is calculated by mapping matching algorithm;Then, the to-be-evaluated index is filtered according to the score, and the corresponding template is called, and finally, the observation and evaluation text report is generated by fusing the template, multi-modal features and infant historical evaluation data.The application can realize the automation, standardization, scientization and safety of infant behavior observation and evaluation, reduce the work burden of teachers, anchor the professional education standard for evaluation, and provide a scientific basis for precise teaching of preschool education.

The application discloses a kind of grain quality, safety and storage risk automatic scoring service implementation method.The method is input with multi-source index, first adopt decimal scaling standardization unified dimension, and obtain characteristic vector by independent component analysis (ICA) and time sequence derivation construction;Subsequently, two layers of random forest integration are constructed: the first layer is trained on bootstrap sampling data by multiple random forest individuals and outputs probability / score, and the performance of the individuals is screened out by joint optimization of feature subset and hyperparameter using artificial bee colony algorithm (ABC) combined with K-fold cross-validation;The output of the retained individuals is spliced with the key original / derived index vector before learning in the second layer, and finally the original grain quality score (QG) / safety risk score (SR) / storage risk score (WR) and comprehensive score (CS) are given.The application improves generalization and robustness while unifying data caliber, and the scoring results can be directly used for business decisions such as collection and storage release, re-inspection interception and storage disposal.

Embodiments of the present description provide an anonymous identity verification method and an apparatus. According to the method, a first device generates a binary indicator vector and an auxiliary vector having a predetermined relationship with the indicator vector on the basis of a sequence number index of a first account in a target account set; and generates a signature value C on the basis of the indicator vector, the auxiliary vector, and a private key of the first account. The first device further calculates a first public key vector on the basis of a first random number selected by a verification device and a public key sequence of the target account set, and calculates a blinded value on the basis of the first public key vector and a random number array selected by the first device. The first device further generates first and second sets of commitment values on the basis of the indicator vector, the auxiliary vector, and a challenge array and a variable value selected by the verification device. The first device provides the signature value C, the blinded value, and the first and second sets of commitment values to the verification device, and the verification device verifies accordingly whether the first account belongs to the target account set.

This invention discloses a language model defense system and method based on dual-channel parallel review and active arbitration, belonging to the field of artificial intelligence security technology. The system includes a pre-filtering layer, a model management unit, a dual-channel review layer, a dynamic arbitration layer, and a response generation module. The method first performs a rapid ternary classification of user input; if deemed suspicious, it simultaneously activates an adaptive intent deep parsing channel and a robust behavior disturbance monitoring channel, outputting intent risk labels and behavior indicator vectors respectively; the dynamic arbitration layer integrates multi-dimensional review results and makes a final interception or release decision through an adaptive learning model. This invention, through layered streaming processing and multi-channel parallel arbitration, effectively defends against advanced jailbreaking techniques such as mimicry attacks, optimization attacks, and automated semantic attacks, significantly reducing false positives and latency while ensuring a high interception rate, thus improving the robustness and real-time performance of the LLM security protection system.