| Summary: | PROJECT RESULTS: Describe the results of your research in reference to its original and/or modified Project objectives. The maximum length for this section is 5 pages (Arial or Verdana font, size 10). During the four years, the project produced 13 WoS papers, 7 international conference papers, and one international conference abstract. Additionally, one completed master thesis, 1 completed PhD thesis (defense during July 2022) and 1 ongoing Ph.D. thesis. In what follows, I will give a description of these products concerning scientific work associated with the eight specific objectives (SO) of the project, thesis guidance, international collaboration, and general public outreach. 1 Scientific work 1.1 SO 1: Develop a learning strategy for the incremental version of TAN that may lead to a forest structure for continuous attributes. In [1] we developed an incremental tree construction procedure for a tree-like continuous Bayesian network classifier. We apply this method to the facial biotype classification problem, an important stage during orthodontic treatment planning. Implementations in R are given. In particular, we developed an alternative learning procedure for the TAN classifier, which we call incremental tree construction augmented naive Bayes (ITCAN). One of the limitations of the TAN model, is that the resulting structure will always be a tree, even if some edges have very low weights (conditional mutual information). With ITCAN, we identify partial TAN solutions where some nodes (attributes) might end up with only the incoming edge from the class. The ITCAN learning procedure with a training set is as follows: 1. Evaluate the accuracy of a naive Bayes classifier using k-fold cross validation. Let this value be Anb. 2. Learn the TAN tree structure. 3. Create a list with the edges in a descending order with respect to their weight (edgeh for h = 1, . . . , n − 1). 4. Assign model naive Bayes model 5. For each h in the list: (a) model model + edgeh (b) Evaluate the accuracy of model classifier using k-fold cross ...
|
|---|