Many datasets include mixed data in them (cuantitative and cualtitative). Thus, a data transformation is needed to train models like support vectorial machines (SVM), artifitial neural networks (ANN) or k nearest networks (k-NN).
In this repo a pattern combinatory logic technique is implemented in several databases through a votation algorithm (AlVot).
You can consult the results of the algorithm in the report(spanish) added to the repo.
The data used to test this algorithm comes from the University of California Irvine (UCI) repository. The datasets are:
- Heart Disease
- Zoo
- Insurance Company Benchmark (COIL 2000)
- Teaching Assistant Evaluation
- HCV data
The full documentation and instructions to run this project will be available soon. You can consult the pdf file (spanish) to review the results of the algorithm.