6results about How to "Low mismatch rate" patented technology

The application relates to the field of artificial intelligence and discloses a crowd advertisement accurate delivery determination method based on a knowledge graph, aiming to solve the mismatch problem caused by the interest model lag of traditional advertisement delivery. The method comprises the following steps: constructing a dynamic evolution type user interest knowledge graph, fusing multi-source behavior flow and static portrait data, introducing a time decay factor and an event triggering mechanism to realize adaptive updating of node weights; through time sequence attention propagation and a cross-domain associated edge dynamic generation algorithm, short-term interest pulses are accurately captured; a double-channel graph neural network inference engine is constructed to process long-term stable and short-term burst interest paths respectively, and a delivery decision is generated through confidence weighted fusion. The application improves the timeliness and accuracy of advertisement matching, reduces the mismatch rate, and enhances user experience and self-calibration capability.

The invention relates to the technical field of engineering digital management and intelligent scheduling, in particular to a dynamic allocation method and device for engineering resources. The method comprises the following steps: acquiring multi-source heterogeneous physical data of a full life cycle of a project, and preprocessing the multi-source heterogeneous physical data; based on the preprocessed physical data, mining a nonlinear relationship between environmental physical parameters and resource consumption by using a gradient boosting decision tree, and predicting a physical resource demand correction coefficient and a working condition risk level of each construction node; and outputting an optimal resource allocation vector on the premise that physical constraints are met by adopting a He-Marx optimization algorithm, a physical resource demand correction coefficient and a working condition risk level and aiming at minimizing physical supply and demand deviation, system operation energy consumption and safety out-of-limit penalty. The method can solve the problems that in the prior art, static planning is difficult to match with real-time changing field requirements, and a general algorithm is difficult to process complex physical constraints, so that resource distribution is not uniform.

This invention relates to the interdisciplinary field of molecular biology and bioinformatics, specifically to a method and system for designing multi-site mutation primers based on rule constraints and multi-objective optimization. The method includes the following steps: receiving the target DNA sequence and mutation site information; automatically grouping the mutation sites according to their position distribution and spacing in the DNA sequence; determining the corresponding primer design region by extending a fixed length upstream and downstream from the center position of each mutation group; generating multiple candidate primer sequences based on the primer design region using a sliding window method; constructing a multi-objective optimization scoring function based on annealing temperature, GC content, and primer length to comprehensively score and screen the candidate primers to obtain the optimal primer; optimizing the 3' end of the optimal primer to preferentially contain stable bases; and outputting the optimized primer sequence and related design parameters and explanation information, effectively solving the pain points of existing multi-site mutation primer designs.

The invention provides a resource allocation method, relates to the technical field of artificial intelligence, and can be applied to the field of financial science and technology. The resource allocation method comprises the steps that in response to received project demand data of a target project, N to-be-allocated resources are obtained by calculating a resource matching degree, and N is a positive integer; and performing conflict detection on the N to-be-allocated resources to obtain L to-be-allocated resources, L being a positive integer.

The invention discloses a voice annotation method and device, computer equipment and a storage medium, belongs to the technical field of artificial intelligence, and is applied to voice data annotation of science and technology finance or health medical scenes. According to the method, a speech overlap detection, speaker recognition and sound source separation cooperative processing mechanism is introduced into a speech annotation process, and overlap detection is firstly utilized to distinguish overlapped segments and non-overlapped segments, so that recognition confusion caused when multiple persons speak at the same time is avoided; a speaker set is constructed based on the non-overlapping segments, speaker prior information is provided for sound source separation of the overlapping segments, and the separation quality and speaker consistency are improved. According to the method, automatic voice recognition and text splicing alignment are performed on each separated sound source branch and non-overlapping segments, so that the manual annotation and segmentation workload can be reduced while the continuity of a time axis is ensured, a uniform voice annotation result is generated, automatic and high-precision annotation of multi-speaker and strong-overlapping voice data is realized, and the voice annotation efficiency is improved. And the voice annotation efficiency and robustness are improved.

The invention discloses a ship pipe identification and assembly method based on a binocular camera. The method comprises the following steps: selecting an optimal acquisition angle to acquire a three-dimensional image pair of a pipe through the binocular camera according to the geometrical shape of the pipe; performing three-dimensional correction on the acquired three-dimensional image pair, generating a depth map, reconstructing a three-dimensional point cloud of the pipe, and extracting geometric parameters of the pipe from the three-dimensional point cloud; taking the stereo image pair, the depth map and the geometric parameters as input parameters of a deep neural network model, and identifying the type of the pipe and a local component of the pipe; performing multi-stage matching on the recognition result and a design database to obtain complete part information; and according to the tube type identification result, the tube matching result and the inventory state, obtaining an optimal set distribution scheme through a reinforcement learning model. According to the method, through fusion of binocular vision and deep learning, the recognition precision is remarkably improved, the second processing speed is greatly improved, the set distribution efficiency is remarkably improved, and the robustness and recognition are enhanced.