Opencl implementation of K-Means Clustering.
How to run?
-
Be sure that you have got proper gcc compiler and linker settings in the visual studio project.
-
Now, you must add src and lib folder from this repository.
-
Create folder's named as "Results" where the clustering results will be stored and "data" where 2D data samples will be stored.
-
i) The dataset(A sets-A1) which is used for project is taken from this webpage http://cs.joensuu.fi/sipu/datasets/. ii) The dataset is two dimensional and has total 3000 samples which is stored as dataset.csv file in "data" folder.
-
Update the macros Num_Cluster, N_Samples and Dimension in the kmeans.cpp. And also, update samples and dimension in kmeans.cl (if dataset is changed).
-
Check correctness of all paths in kmeans.cpp (for dataset.csv, kmeans.cl and results respectively.)
-
Build and run the program and you will be able to observe the performance and result which is stored in .csv file in "Results" folder.