| In yesterday's compiler seminar, I found that ------ --- and -------- ---- have | |
| this data mining program that could be useful for the paper. It is very compute | |
| intensive and creates one large tree and then repeatedly creates a small tree, | |
| uses it for a search, and then throws it away. That latter phase takes 96% of | |
| the total run time. The code is in C++. | |