.png){kind=icon}
.png){kind=icon}
Return
HMP: Efficient heuristic algorithms for MDL-based itemset mining
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2025.114314.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Introduces HMP, a framework combining Simulated Annealing and Hill Climbing with the Minimum Description Length principle for efficient pattern selection. • HMP enhances computational efficiency and pattern quality, reducing runtime while ensuring representative itemsets. • Demonstrates superior performance over Grimp, Krimp, and Slim in runtime and compression ratio, and excellent downstream classification accuracy.