12.07.2015 Views

Predicting Cardiovascular Risks using Pattern Recognition and Data ...

Predicting Cardiovascular Risks using Pattern Recognition and Data ...

Predicting Cardiovascular Risks using Pattern Recognition and Data ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Adding attribute weights to the clustering algorithm does not change the computational cost of thealgorithm compared to the KMIX. The WKMIX results (Table 8.2) clearly show an improvement inaccuracy compared to KMIX. Furthermore, the high performances (ideal rates) in Table 8.3 show thesupervised techniques can perfectly replicate the outcomes of the WKMIX clustering models. Thismeans the data is formed into well defined clusters by the WKMIX algorithm. This suggests that byadding weights in the clustering process, the quality of the algorithm is improved.8.7. SummaryThis chapter presented the concept of feature selection. A simple (filter) feature selection method isapplied to the thesis data to measure the relevant level of attribute set to the outcomes. The improvedWKMIX algorithm provides an investigation of <strong>using</strong> attribute weights (mutual informationcalculations) in the clustering process. The Case Study VI results proved this by showing the increasingperformance compared to the KMIX.138

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!