Menu
Index

Machine learning

The aim in the module is to classify, cluster, or find the relationship using the desired features taken from the signals. Therefore, popular machine learning algorithms existing in this module are used.
The parameters each machine learning algorithm has can be changed from the components on the panel or from the options that appear during the study.
The module can store each analysis table, learning model and its results by creating a special name. It also can compare these results statistically. Then, the results are written into a standard ASCII file with space delimiters, a format compatible for importation of the results into most spreadsheets and statistical software programs (i.e. EXCEL, SPSS, SYSTAT, etc.).The learning models stored are intended to be used in a real-time system. In addition, when it is intended to use the data in another software except the module, the data can be exported into file formats such as ARFF  so that they can be run by another software or they can be imported to the database from these file formats. Futhermore, when it is intended to use the data in another software except the module, the data can be exported into file formats such as ARFF  so that they can be run by another software or they can be imported to the database from these file formats.
The program also gives all the machine learning tables in the chosen database in turn to WEKA software as input data. The module can store the results as an exclusive file for the analysis when desired and can make comparisons by running the algorithms as well as the parameters available in WEKA as a batch processing.
 
The online help was made with Dr.Explain