GraphConstruction

For any given dataset nd , where n is the number of examples and d is the dimension of an example, we need to construct the Graph g=(V,E) where |V|=n and E is set of edges (Our focus is construction of "Weight Matrix W of size nn" , each entry in W, W_ij, denotes weight between node i and node j). In simple terms

Input : Data Matrix nd output : Weight Matrix nn

Process : (a) Method for choosing how many edges needs to be connected to a node like k-NN and e-NN (b) Distance Measure to compute distance between node i and node j, namely Euclidean distance, Gaussian Kernel,Cosine Similarity, etc.

Usage :

1. param.config file used to set all the required parameters.

To Run

1. Run Makefile $> make
2. Makefile will generate executable hello,to run the program pass param.config as input to executable $> ./hello param.config.
3. Wait till program finishes, according to param.config, the ouput Weight Matrix n*n is written to 'weight.txt' in same folder.