الفهرس 
Towards an Effective Approach to Detect the Hidden Patterns
AbstractOnly 14 pages are availabe for public view
 |  | from | 145 |  |  |
| | | from | 145 | | |
Abstract
Data mining is considered to be a field of research whose objective is to investigate good knowledge from large amount of data. Data mining field has been widely spread in the past two decades in computer science area. One of the most widely spread data mining technique is K- Nearest Neighbour, because it is simply used and can be applied on different types of data. The presence of multidimensional and outliers data have great effect on the accuracy of the KNN algorithm. In this thesis, a new hybrid approach called Particle Optimized Scored K-Nearest Neighbour was proposed in order to improve the performance of KNN. The new approach is implemented in two phases; the first phase helps to solve the multidimensional data by making feature selection using Modified Particle Swarm Optimization algorithm, the second phase helps to solve the presence of outliers by taking the result of the first phase and apply on it a new proposed scored KNN technique. The proposed approach was applied on two datasets; first, Soybean dataset which is composed of 36 attributes, 15 classes and 684 instances. Second, real life dataset gathered from 3 dental clinics in El Baharia Oasis in Egypt which is composed of 56 attributes, 2 classes and 328 instances. The model was implemented using 10 fold cross validation. Experiments showed that the proposed hybrid approach gave better performance than classical KNN and Modified KNN described in the related workReem Kadry and Osama Ismael, 2A New Hybrid KNN Classification Approach based on Particle Swarm Optimization3, International Journal of Advanced Computer Science and Applications (IJACSA), vol. 11, Issue 11, 2020. The journal is considered to be indexed, Scopus and web of science