Abstract
In order to address the problem that high energy consumption, high memory usage and low clustering effect in traditional data set high-dimensional clustering algorithms, we propose the high-dimensional clustering algorithm of incomplete mixed data set based on artificial intelligence. First, we construct the phase space reconstruction to ensure the invariance of features of incomplete mixed data set by analyzing the incomplete mixed data set and introduce the correlation dimension to obtain the feature correlation value. Second, we introduce the standard deviation and realize the extraction of features of incomplete mixed data set through calculating the sparsity of sample features. Third, we conduct repeated clustering for the mixed data set in the subspace according to the degree of correlation between incomplete mixed data sets in the multidimensional subspace. Last, we realize the design of high-dimensional clustering method for incomplete mixed data set in accordance with the stronger relevance in the mixed data sets. Experimental results show that the proposed algorithm has good correlation dimension processing effects, lower memory usage, time-consuming, lower and concentrated ensemble energy consumption (within 300J), good clustering effects, as high as 92%, which has some advantages and practical application value.
Highlights
Artificial intelligence covers a wide range of fields, including computer vision, machine learning, with the main objective to complete the tasks with machine [1], [2] that can only be completed with human intelligence, which has higher requirements for complexity and time [3]
Through comparing the clustering performance of image data sets and clusters generated by synthesis and verifying the clustering performance of the method projecting on the random line based on data, we get the conclusion according to probability density of measure: As the structures we found in the image data set do not meet the traditional clustering concept, we further propose a rapid method for high-dimensional data clustering with a layered approach
To verify the comprehensive effectiveness of highdimensional clustering algorithm of incomplete mixed data set based on Artificial intelligence, we need to conduct many experimental tests
Summary
Artificial intelligence covers a wide range of fields, including computer vision, machine learning, with the main objective to complete the tasks with machine [1], [2] that can only be completed with human intelligence, which has higher requirements for complexity and time [3]. The field of artificial intelligence usually involves data clustering and data classification. Classification is divides incomplete data with mixed values into existing categories according to the characteristics or attributes of data. Clustering is to divide highsimilar things into meaningful and useful data group clusters according to the similarity of things [4], [5]. As one of the computing methods of Artificial intelligence, clustering analysis collects [6]–[8].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
