Optimization theory-based parallel clustering method with scale constraint

Abstract

The invention relates to an optimization theory-based parallel clustering method with scale constraint, which comprises the following steps of: S1, acquiring a data set to be clustered and a scale constraint vector, and initializing parameters according to the data set and the scale constraint vector; S2, decomposing the data set subjected to parameter initialization into a plurality of sub-problems through a distribution matrix; S3, introducing a Lagrange multiplier vector, solving and clustering the sub-problems in parallel through a projection matrix, and updating the allocation matrix according to solving results of the sub-problems; and S4, calculating a convergence judgment parameter, judging whether clustering reaches a convergence stopping criterion or not according to the convergence judgment parameter, stopping iteration and outputting the current allocation matrix and the corresponding clustering result if the convergence stopping criterion is met, and otherwise, continuingto execute the steps S1-S3 for iteration. Compared with the prior art, the method has the advantages that a large-scale data set is processed in finite iterations, and a high-quality clustering resultis provided.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

This technology improves cluster analysis efficiency for big data applications without adding any extra constraint on its dimensions. It uses an iterative method called LAGLE (Lagrangians Multiplicator Vector) which reduces dimensionality issues while optimizing parameters simultaneously. Additionally, it achieves full parallelization over multiple machines efficiently even when dealing with very small datasets. Overall, this technique makes efficient use of computational resources more effective than previous methods.

Problems solved by technology

This patented technical solution discussing how clumped together two or three datasets may help improve performance for various applications like big data analytical processing tasks. These techniques involve dividing up larger sets of data into small subsets called chunks, allowing multiple instances of these subsets to work equally well without overlapping their content. Additionally, they require specific scaling restrictions at any scales where the dataset needs to match its own properties. Existing approaches either lack sufficient accuracy or cannot handle significant differences across all dimensions accurately enough.

Images