ClassBasedAppriori is an complex association mining algorithm created as ad-hoc development* for the company ELİAR Elektronik San. A.Ş., Automation Solutions for Textile and Glass Batch Industries.
ad-hoc development: creating software without any formal guidelines or processes
- It is possible to mine associations in data with univariate and mutivariate classes.
- It is possible to mine associations in continuous data by discretizing it.
- It is possible to mine associations with preselected conditions for root cause analysis.
- ClassBasedApriori scans the dataset only twice thanks to divide and conquer search method, where Apriori scans multiple times using breadth-first search method.
- ClassBasedApriori finds all possible associations and can run an intelligent algorithm on the associations for data mining.
- ClassBasedApriori allows user to draw associations within and between different classes.
- ClassBasedApriori allows user to find associations with certain conditions in a supervised manner.
- Since the computational power requirement increases exponentially with the size of the data, it is difficult to work with a large database.
- It processes all necessary or unnecessary information, there is no supervision over the information to be processed.
- On-demand data mining, disabling the brute-force approach
- Apply parallel processing to improve performance
- Interactive plotting to enable exploratory data analysis
- Option to dynamically change support, confidence and lift thresholds while plotting
- Intelligent relative association detection
- Intelligent association filtering
