This paper presents a semi-supervised kernel-based fuzzy c-means algorithm called S2KFCM by introducing semi-supervised learning technique and the kernel method simultaneously into...
Currently the most accurate WLAN positioning systems are based on the fingerprinting approach, where a “radio map” is constructed by modeling how the signal strength measureme...
In this paper we present some lessons learned from building vizsla, the keyword search and topic classification system used on the largest Hungarian portal, [origo.hu]. Based on ...
This paper presents a semi-supervised learning (SSL) approach to find similarities of images using statistics of local matches. SSL algorithms are well known for leveraging a larg...
Learning Bayesian Belief Networks (BBN) from corpora and incorporating the extracted inferring knowledge with a Support Vector Machines (SVM) classifier has been applied to charac...