MATLAB codes provided in this document are tested on 2019b and 2018b MATLAB versions. MinGW is a supported C/C++ compiler, which is required to run some of the C++ dependencies. Additionally Linspecer from MathWorks is used to generate attractive colour combinations and shades for beautiful visualisations.
The complete procedure to operate the GUI consist of four stages.
At first, the user has the option to select any clustering algorithm from the drop down list consisting of ten aforementioned algorithms to classify the data. In the second stage, user requires to input data that needs to cluster.
There are three different options are available in which user can enter data depending on the nature of data available.
User can input raw data, spikes and in form of extracted features.
If the inputted data is in the form of raw data or spikes data then user also has to select either PCA or Harr Wavelets feature extraction technique from drop down list to extract data features for classification.
The last step is to select the N number of groups the data should be divided to run the GUI.
The value of N depends on the length of datasets and type of clustering algorithm referred to the time analysis section of this paper.
Press “Start Clustering” to start the clustering process.
The GUI yields the clustering labels with high accuracy and in a fast and efficient way.
The first graph in the GUI shows the clustered spikes and the second graph illustrates the clustered features of the inputted data.