A model for early prediction of Alzheimer disease using a hybrid data mining
technique 
   Lead Investigator:    Alaa Bessadok
   Institution      :    High institute of management - Gabes , Tunisia
   E-Mail           :    alaa.bess@gmail.com
   Proposal ID      :    892



   Proposal Description:
   In a research project on my master degree, I want to apply different models on a NACC dataset in order to extract relevant knowledge that will help the early prediction of the Alzheimer disease that will be applied on new cases. The model consists of a hybridization of two data mining techniques and precisely a classification and an association algorithm.  I choose the SVM, Artificial Neural Network and AdaBoost as classification algorithms which i want to apply on the data, in favor of making a comparison between them based on some measures such as accuracy, sensitivity and specificity. The Frequent Pattern Tree algorihtm, the association technique I chose, will be used to find a correlation between data in the dataset that may be useful for doctors and researchers in the medical field.  So, the goal of my contribution in this project is: make some informations related to a specific person as an input of the classifier, and predict the risk's level of being affected by alzheimer's disease (low, medium and high risk) which is based on the significant pattern obtained from the FP-tree algorithm.