An efficient feature selection algorithm based on the description vector and hypergraph

The curse of dimensionality is a bottleneck in big data and artificial intelligence. To reduce the dimensionality of data using the minimal vertex covers of graphs, a discernibility matrix can be applied to construct a hypergraph. However, constructing a hypergraph using a discernibility matrix is a time-consuming and memory-consuming task. To solve this problem, we propose a more efficient approach to graph construction based on a description vector. We develop a graph -based heuristic algorithm for feature selection, named the graph-based description vector (GDV) algorithm, which is designed for fast search and has lower time and space complexities than four existing representative algorithms. Numerical experiments have shown that, compared with these four algorithms, the average running time of the GDV algorithm is reduced by a factor of 36.81 to 271.54, while the classification accuracy is maintained at the same level.

Automated summary

The curse of dimensionality is a bottleneck in big data and artificial intelligence. To address this issue, a more efficient approach to graph construction based on a description vector is proposed. The graph-based description vector (GDV) algorithm is developed for fast search and has lower time and space complexities than four existing algorithms, while maintaining the same level of classification accuracy.