MINING INTERACTION PATTERNS BY CLUSTERING

Author’s Name : Madhumathi D | Meghana S | Rashmi P | Sanjana R |L Babitha | R Dinesh Kumar

Volume 02 Issue 02  Year 2015  ISSN No:  2349-3828  Page no: 10-14

Abstract – Human brain is complex and its difficult to understand its activities. fMRI (Functional Magnetic Resonance Imaging) provides the details to study about the brain functions fMRI  process requires the effective and efficient data mining techniques to obtain the data. To collect the complex interaction patterns among brain a special type of clustering technique is required. The objects with the similar interaction pattern are to be assigned in the similar cluster. Based on this process an Interaction K-Means clustering algorithm is proposed to perform the clustering of objects in the training set of data. By the proceeding cleavage the percepts are obtained. The psychiatric disorders can be effectively recognized by means of the Artificial neural networks by which the results can be obtained effectually.

Keywords – fMRI; Data Mining; Cluster; Neural Networks