Second Generation of Mass Estimation

Abstract

We made three progresses in the field through mass estimation: First, we propose the first adaptive version of mass estimation using a new nearest neighbor procedure which runs significantly faster than existing nearest neighbor procedures, and it needs no indexing schemes. Second, we propose the first mass-based Bayesian classifier which estimates the likelihood directly in multi-dimensional space; unlike existing Bayesian classifiers which estimate simplified surrogates of likelihood (e.g., one-dimensional likelihood). Third, we have created the first mass-based similarity measure and show that it is an effective alternative to distance-based similarity measure in content-based information retrieval problems. 1 Report Documentation PageForm Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. REPORT DATE SEP 20132. REPORT TYPE PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)Gippsland School of Information Technology, ,Monash University,Victoria,Australia,AU,3842 PERFORMING ORGANIZATION REPORT NUMBER N/A SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) AOARD, UNIT 45002, APO, AP, 96338-5002 SPONSOR/MONITOR'S ACRONYM(S) AOARD SPONSOR/MONITOR'S REPORT NUMBER(S) AOARD-114112 DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited SUPPLEMENTARY NOTES ABSTRACTThree progresses were made in the field through mass estimation: 1) the first adaptive version of mass estimation using a new nearest neighbor procedure which runs significantly faster than existing nearest neighbor procedures and needs no indexing schemes, 2) the first mass-based Bayesian classifier which estimates the likelihood directly in multi-dimensional space; unlike existing Bayesian classifiers which estimate simplified surrogates of likelihood (e.g., one-dimensional likelihood), and 3) the first mass-based similarity measure which can be an effective alternative to distance-based similarity measure. SUBJECT TERMS Information diffusion, Opinion formation, Social network, Burst detection, Influential nodes, Network dynamics, Knowledge discovery from network SECURITY CLASSIFICATION OF:17. LIMITATION OF ABSTRACT Prescribed by ANSI Std Z39-18 In addition, we have extended the two previous works on mass estimation and published them in Machine Learning Jo...