EFFECTIVE LEARNING OF RESTRAINTS FOR SEMI ADMINISTERED CLUSTERING
Journal Title: Elysium Journal of Engineering Research and Management - Year 2015, Vol 2, Issue 3
Abstract
Semi-supervised clustering is improving clustering performance by considering user supervision in the form of pairwise constraints. We study the active learning problem of selecting pairwise must-link and cannot-link constraints for semi-supervised clustering. We apply a general framework that builds on the concept of neighborhood, where neighborhoods contain “labeled examples” of different clusters according to the pairwise constraints. Our active learning method expands the neighborhoods by selecting informative points and querying their relationship with the neighborhoods. Under this framework, we build on the classic uncertainty-based principle and present a novel approach for computing the uncertainty associated with each data point.
Authors and Affiliations
Balaji E. , Lavakumar M. G.
Keywords
Related Articles
- EP ID EP386433
- DOI -
- Views 112
- Downloads 0
HIGH SPEED AND LOW POWER CMOS TECHNOLOGY BASED RAM-CAM MEMORY DESIGN
ontent-Addressable Memory (CAM) is a special type of memory, which is used in very-high speed searching applications. The other names of CAM are the associative memory, associative storage and the assoc...
THE CONCEPTUAL VIEW OF TESTING AS A SERVICE (TAAS) ON CLOUD
Cloud computing leads a chance in contributing a Testing as a Service (TaaS) for cloud, and cloud-based applications. It offers new business challenges, opportunities, and demands in advanced service mo...
HIGHLY ACCURATE RECOGNITION OF BODY MASS INDEX AND HUMAN DIABETICS THROUGH SENSORS
Monitoring of diabetics and BMI (Body Mass Index) is used in many medical field and research applications. This paper suggests a method for performing highly accurate recognition of pressure, sugar, he...
DEVELOPMENT OF ERROR CORRECTION MECHANISM BASED ON RCIC TURBO CODES IN LTE NETWORK
This paper describes an error correction mechanism based on Rate-Compatible Insertion Convolutional (RCIC) Turbo codes.Rate-compatible Turbo codes are part of numerous mobile communication systems (...
AN EFFICIENT DAUBECHIES COMPLEX WAVELET BASED MULTI-RESOLUTION APPROACH FOR MULTIMODAL MEDICAL IMAGE FUSION
Multimodality approach for medical images provides multiple advantages for the detection, diagnosis and management of the diseases. Image fusion techniques are used to combine the high resolution images with...