Design and Application of Multi Dimensional Data Clustering Analysis Platform based on the Combination of Deep Learning and K-means Algorithm

In order to solve the common problems of dimensionality curse and insufficient clustering accuracy in multidimensional data analysis, this paper proposes a multidimensional data clustering analysis method that combines deep learning and K-means algorithm to improve the accuracy and efficiency of clustering analysis. First, a deep neural network is used to extract features from the input multidimensional data to generate potential feature representations in high-dimensional space. Through multi-layer nonlinear transformations, the deep learning model can capture complex patterns and relationships from the original data and effectively reduce the noise impact of the original data. Then, the K-means algorithm is used to perform clustering analysis on the extracted features. Specifically, K-means divides the data into different clusters by minimizing the intra-class variance. In this process, the appropriate number of clusters k and initialization strategy are selected to avoid falling into the local optimal solution. Finally, the clustering results are evaluated using indicators such as the silhouette coefficient to further optimize the clustering accuracy and ensure the effectiveness and stability of the classification results. From the perspective of the silhouette coefficient, the silhouette coefficient of conventional K-means clustering is generally between 0.39 and 0.40, while the result after adaptive clustering optimization is generally around 0.42, which effectively solves the clustering accuracy problem in high-dimensional data analysis.