Application and implication of edge selection
The project explored Edge Selection procedure to fit an undirected graph to given datasets. Undirected graphs are frequently used to represent, model and analyse associative relationships among the entities on a social, biological and genetic networks. Undirected graphs do not have direction. They have edges that do not follow any direction. Nodes are vertices that represent objects and edges are the connections between the objects. In undirected graphs edges represent two-way relationship, where each edge can be traversed in both directions. We explored many cross validation techniques and chose k-fold (k=10) cross validation to reach an accurate predictive model. This method is used to select undirected graphs from various real data sets. We ran it on many in built R datasets as well as datasets on topics ranging from Facebook data, finance data, marketing data, graduate admission data, life expectancy data etc. We also created some new functions for "ES" package in R.