A data visualization tool has been implemented which uses the aforementioned algorithms for data analysis along with further methods. The visualization/analytics tool combines methods like Silhouette and Elbow to retrieve possible number of clusters from a large amount of data, and then it gives the possibility to train an SVM with any chosen kernel according to the number of retrieved clusters. Moreover, it offers the ability to use an innovative method implemented for the linear trend analysis of timeseries data. This knowledge discovery application has been designed in order to combine the best methods and algorithms applied to real data of an industrial pilot (CERTH/CPERI) that gave the most useful information. It is flexible so that it could be used with minor adaptations to similar problems and cases in the industrial area.
| Web resources: | https://www.youtube.com/watch?v=00dN6v1RjtI |
- No items published
- Not specified