Show description

3 CF-KNNC and KD-KNNC Algorithms We also propose a combination of a tree based clustering and KNNC with a view to combine advantages of KNNC, CF-Tree and KD-Tree algorithms. KNNC essentially provides good accuracy, however takes longer to run. 1 CF-KNNC Algorithm This algorithm has two phases, design and classification. Where the design phase constructs level restricted CF-Tree, and in classification phase it runs KNNC on the selected leaves of the CF-Tree. To improve the execution time we added a heuristic where KNNC is avoided if all the patterns in a cluster have the same class label.

Thus, a player stays related with his closer neighbor belonging to his zone and the closer neighbor belonging to the neighbor zone. In this work, the goal is to build a simple but robust structure based on relevant relations able to manage the dynamic nature of the game in a topological way, based on triangular sub-graphs that are built as indicated below: – Step 1. Let A = {ai < bi < ci < ... < ni } be finite number of nodes (players) belonging to a zone Zi . Where player ai is located at the top of the zone and player ni at the bottom of it.

Proceedings of the 1996 ACM SIGMOD International Conference on Management of Data, Montreal, Quebec, Canada, pp. mx Abstract. This work is focused on the recognition of team patterns represented by different formations played by a soccer team during a match. In the soccer domain, the recognition of formation patterns is difficult due to the dynamic and real time conditions of the environment as well as the multiple interactions among team mates. In this work, some of these multiple interactions are modeled as relations represented by a topological graph which is able to manage the dynamic changes of structures.