- Jingbo Shang, Wenzhu Tong, Jian Peng, and Jiawei Han, "DPClass: An Effective but Concise Discriminative Patterns-Based Classification Framework", In Proceedings of 2016 SIAM International Conference on Data Mining (SDM 2016), 2016
- g++ 4.8
- matlab
The current executables require OpenMP, which does not come by default on OS X. To be able to run it on OS X, follow this stackoverflow post.
DPClass can be easily built by Makefile in the terminal.
$ make
You could execute the code in the following way:
./run_forward.sh dataset_name l2_norm_coefficient
./run_glmnet.sh dataset_name
Example:
./run_forward.sh adult 0.5
./run_glmnet.sh adult
All the parameters are located in run_forward.sh and run_glmnet.sh. Some important parameters are as follows.
TOPK=20
Top-k discriminative patterns.
MAX_DEPTH=6
The maximum number of conditions in a single pattern.
MIN_SUP=10
Minimum support threshold.
####For other parameters regarding each individual module, please check the corresponding cpp files.