2 results about "Entropy criterion" patented technology

This invention discloses a fire-fighting pipe fitting leakage risk early warning system based on big data analysis, comprising the following steps: collecting multi-dimensional time-series data such as pressure, flow rate, temperature, and humidity; constructing a data processing and modeling workflow; employing kernel canonical correlation analysis to extract nonlinear correlation features between different monitoring parameters to identify weak correlation changes before leakage; and constructing an anomaly measurement mechanism based on the maximum correlation entropy criterion to quantify the degree of feature shift. By fusing the above correlation features and entropy information, a dynamic risk index is generated and compared with a dynamic threshold to achieve real-time early warning of leakage risk, effectively supporting early fault detection and intelligent assessment of fire-fighting pipe fittings. This invention achieves dynamic perception and intelligent judgment of fire-fighting pipe fitting leakage risk, possessing data-driven early warning capabilities.

This invention discloses an adaptive MLP pruning method for large-scale vision Transformers, addressing the practical deployment requirements of real-time visual perception, low power consumption, low latency, and high throughput in scenarios such as robots, drones, and mobile terminals. It solves the technical problems of existing large-scale vision Transformers, including redundant parameters, high computational and memory overhead, slow inference speed, and difficulty in edge deployment. The core method first accurately evaluates the importance of hidden neurons in the MLP based on Taylor expansion combined with the information entropy criterion. Then, it adaptively prunes and sorts the neurons according to the redundancy of different MLP modules using a binary search algorithm. Combined with knowledge distillation, it restores the performance of the pruned model, ultimately achieving a reduction of approximately 40% in parameters and computational load, and increasing the inference speed to about 1.5 times the original. This method not only demonstrates near-consistent performance with the original model in multiple benchmark tests such as zero-shot image classification, retrieval, and kNN evaluation, but also slightly surpasses it in some scenarios. Furthermore, it has the advantages of not relying on the original model's loss function and additional modules, and being compatible with word reduction methods. Ultimately, this invention provides a lightweight, high-performance, and easily adaptable practical solution for various application scenarios in the field of computer vision that require efficient model inference, and powerfully promotes the low-cost engineering deployment and commercialization of large-vision Transformer models.