This kit contains a set of python libraries for finding user interest-based patterns in the data. The python programs in this toolkit belong to one of the following three topics.
- Native sequential algorithms - traditional pattern mining algorithms whose input exists in the form of transactional databases (or files).
- DataFrame sequential algorithms - Native sequential algorithms are modified to accept dataframes as input. These algorithms can be exported to other softwares like Azure.
- Native parallel algorithms - We provide algorithms for finding patterns in Map-Reduce framework.
Procedure for the creation of files:
Pami_ckit/<nativeSequential/dataFrame/nativeParallel>/PatternModel/Algorithm
E.g. pami_ckit/nativeSequential/FrequentPatternMining/FPgrowth