Exploiting empirical variance for data stream classification

Classification, using the decision tree algorithm, is a widely studied problem in data streams. The challenge is when to split a decision node into multiple leaves. Concentration inequalities, that exploit variance information such as Bernstein’s and Bennett’s inequalities, are often substantially strict as compared with Hoeffding’s bound which disregards variance. Many machine learning algorithms for stream classification such as very fast decision tree (VFDT) learner, AdaBoost and support vector machines (SVMs), use the Hoeffding’s bound as a performance guarantee. In this paper, we propose a new algorithm based on the recently proposed empirical Bernstein’s bound to achieve a better probabilistic bound on the accuracy of the decision tree. Experimental results on four synthetic and two real world data sets demonstrate the performance gain of our proposed technique.     

Development of 19-plex Y STR system and polymorphism studies in Pakistani population

For the development of 19-plex Y STR system and polymorphism studies in locl ethnic populations sixteen markers of non-recombining regions (NRY) of Y chromosome, which show high power of discrimination among individuals, were selected in this study. Blood samples (600) were e.ollected from the males of three most common castes of Pakistani population (Arnin, Awan and Rajput) with different parent lineages. Three markers (DYS385a/b, DYS389Ⅰ/Ⅱ and YCAⅡa/b) among 16 Y STRs are double-targeted regions of the Y chromosome and thus provide two polymorphie peaks for each respective primer set. These 16 Y-STRs were developed into Megaplex system for simultaneous amplification of all markers within the population. The overall power of discrimination observed in focused populations was 60.5%, 66.5% and 55% in Rajput, Awan and Arain casts respectively. This discrimination power will be helpful in haman identification for forensic casework studies including sexual assaults and paternity testing.