This branch revamps the cluster library to add a function that searches for a good number of splits based on various provided heuristic functions. It also changes how cluster splitting is iteratively performed to be more robust in the case where the tree might contain more information than the clustered set of instances. This also introduces log_likelihood functions to the node classes that are used in the new heuristic functions.
This branch revamps the cluster library to add a function that searches for a good number of splits based on various provided heuristic functions. It also changes how cluster splitting is iteratively performed to be more robust in the case where the tree might contain more information than the clustered set of instances. This also introduces log_likelihood functions to the node classes that are used in the new heuristic functions.