Andreas Leonidas Prodromidis
RESEARCH INTERESTS
E-commerce, Knowledge Discovery and Data Mining, Distributed Systems, Machine Learning,
Agent Technologies, Information Retrieval.
-
PROFESSIONAL EXPERIENCE
- Director, Research and Development , iPrivacy LLC,
1999-present
- First employee of the company and one of the initial designers of the iPrivacy system, a comprehensive system for anonymizing and securing E-Commerce (private surfing, private shopping, private shipping).
- Designed and developed a Java-based proxy web server.
- Database architect, programmer, and DBA.
- Initiated, designed and prototyped a private voice mail system using
software-based PBX (telephone center) on LINUX.
- Graduate Research Assistant, Columbia University,
1995-1999
Parallel and Distributed Intelligent Systems Laboratory
- Designed, developed and evaluated JAM, a distributed data mining system for managing intelligent learning agents across multiple database sites to gain and combine new knowledge. JAM was used in credit card fraud detection with impressive results. It analyzed real credit card transactions provided by two banks, members of the FSTC, and computed models that explain anomalous or errant behavior. When JAM used these models as fraud predictors on new transactions, it outperformed the standard COTS (Commercial Off The Shelf) detectors employed by the industry. The results were presented in many computer science conferences.
- Graduate Research Assistant, Columbia University,
1993-1995
Mobile Computing Laboratory
- Extended an MPEG video player to operate across mobile links and
adapt to changing network conditions. Introduced a proxy host, capable
of filtering incoming MPEG streams (destined to the mobile hosts) to
accommodate bandwidth changes.
- Modified the SunOs kernel to optimize file and path name lookup,
augmenting the VFS lookup algorithm and the NFS protocol. Conducted
performance tests.
- Supplemental Research Staff, IBM T.J. Watson Research
Center, March 1996 - September 1997
Network Servers department. (Full time during the summer, part time
during the rest of the year).
- Participated in the design and implementation of a prototype system for combining data analysis algorithms with databases on the Internet.
- Participated in the Lava Project, replacing JavaOS's core kernel
with a new microkernel developed at IBM
- Supplemental Research Staff , IBM T.J. Watson Research
Center, May 1995 - March 1996
Cluster and Distributed Systems department. (Full time during the
summer, part time during the rest of the year).
- Developed and evaluated OCTOPUS on a multi-node IBM SP2 machine running AIX 4.1. OCTOPUS is an extensible, hierarchical, reconfigurable and fault tolerant resource management system for
cluster computing environments.
- Software Engineer , Intrasoft Corp., 1992-1993
- Developed the Systems Hardware Architecture Presentation Environment
(SHAPE) using the Tool Builders Kit in X-Windows systems. SHAPE is a
software system that supports the architecture design of real time
systems. The project was part of the RACE programs, funded by the
European Union.
EDUCATION
- Ph.D., Computer Science Department, Columbia University,
October 1999
Advisor: Professor Salvatore J. Stolfo
Thesis: Management of Intelligent Learning Agents in Distributed
Data Mining Systems
- M.Phil. Computer Science, Columbia University,
October 1996
Advisor: Professor Salvatore J. Stolfo (GPA: 4.0)
- M.S. Computer Science, Columbia University,
October 1994
Advisor: Professor Daniel J. Duchamp (GPA:4.0)
- Diploma Electrical and Computer Engineering, National
Technical University of Athens, July 1993
Advisor: Professor Emmanuel Protonotarios (GPA: 9.34/10.0)
Thesis: Methodology and Tools for the Automated Description of
Telecommunication Systems.
-
HONORS AND AWARDS
- IBM Corporate Fellowship, 1996-1998
- Runner-Up Best Paper Award, KDD'97, "JAM: Java Agents for
Meta-Learning", 1997
- Kollias Award and Fellowship, 1992
- Merit Scholarships by Greek National Scholarship Foundation
(5 times), 1989-1993
- Distinction in the Greek Mathematics Olympics, 1987
-
DISSERTATION
Management of Intelligent Learning Agents in Distributed Data Mining Systems
Data mining systems aim to discover patterns and extract useful information
from facts recorded in databases. In this dissertation, we concentrated on the
problem of managing a distributed collection of intelligent learning agents
that compute and combine descriptive models over large and distributed
databases. I designed and implemented JAM (Java agents for
Meta-learning), an extensible agent-based distributed data mining system that
supports the remote dispatch and exchange of agents among participating data
sites and employs meta-learning techniques to combine the multiple models that
are learned. JAM is a scalable system, designed to make effective use of the
available system resources and implemented to operate across heterogeneous
platforms. Furthermore, it is adaptive to changes in its computational
environment and is flexible to accommodate databases with differences in their
schema definitions. An empirical analysis of JAM on the real-world data mining
task of fraud detection in financial information systems, evaluated the
effectiveness and performance of our approaches and demonstrated their
potential utility.
TEACHING EXPERIENCE
- Project Student Supervisor, Columbia University,
Spring 1996 - present
Supervised several Masters and Undergraduate project students to enhance and extend JAM.
- Instructor , Columbia University, Fall 1996
Introduction to the C Programming Language (W3101)
- Teaching Assistant, Columbia University, Spring 1994 -
Spring 1996
Computability and models of Computation (W3261), Computer Organization
(W3824), Programming Languages and Translators (W4115), Artificial
Intelligence (W4701)
- Teaching Assistant, National Technical University of
Athens, Spring 1993
Introduction to Databases
-
PUBLICATIONS
-
Refereed Journal Publications
Andreas L. Prodromidis and Salvatore J. Stolfo, "Cost Complexity-based
Pruning of Ensemble Classifiers", Knowledge and Information Systems
(KAIS) (accepted, in press) 2001.
Philip K. Chan, Wei Fan, Andreas L. Prodromidis and Salvatore J. Stolfo,
"Distributed Data Mining in Credit Card Fraud Detection", IEEE Computer
Society, December 1999.
-
Refereed Book Chapters
Andreas L. Prodromidis, Philip Chan and Salvatore J. Stolfo,
"Meta-Learning in Distributed Data Mining Systems: Issues and
Approaches", Book on "Advances of Distributed Data Mining",
editors Hillol Kargupta and Philip Chan, AAAI press, 2000.
Nayeem Islam, Andreas L. Prodromidis and Mark Squillante. "Dynamic
Partitioning in Different Distributed Memory Environments", (Long
Version of the IEEE IPPS/SPDP'96 paper). In Job Scheduling Strategies
for Parallel Processing, pages 244-270, Lecture Notes in Computer
Science Vol. 1162 (Springer-Verlag 1996)
-
Refereed Conference Publications
Salvatore J. Stolfo, W. Fan, Wenke Lee, Andreas L. Prodromidis and
Philip K. Chan, "Cost-based Modeling for Fraud and Intrusion
Detection: Results form the JAM Project", Proceedings of DARPA
Information Survivability Conference and Exposition, IEEE
Computer Press, 2000.
Andreas L. Prodromidis and Salvatore J. Stolfo, "Minimal Cost
Complexity Pruning of Meta-Classifiers", Sixteenth National
Conference on Artificial Intelligence, AAAI-99, Orlando, Florida,
July 1999.
Andreas L. Prodromidis and Salvatore J. Stolfo. "Pruning
Meta-Classifiers in a Distributed Data Mining System", In Proc. of the
First National Conference on New Information Technologies (NIT'98).
Athens, Greece, October 1998.
Andreas L. Prodromidis and Salvatore J. Stolfo. "Mining databases
with different schemas: Integrating incompatible classifiers", In Proc.
of the Fourth AAAI International Conference on Knowledge Discovery and
Data Mining (KDD'98). New York City, New York, August 1998.
Salvatore Stolfo, Andreas L. Prodromidis, Shelley Tselepis, Wenke Lee,
Wei Fan and Philip Chan. "JAM: Java Agent for Meta Learning Over
Distributed Databases", In Proc. of the Third AAAI International
Conference on Knowledge Discovery and Data Mining (KDD'97). Newport
Beach, California, August 1997 (Runner-up best applications
paper).
Nayeem Islam, Andreas L. Prodromidis, Mark Squillante, Liana Fong and
Ajei Gopal. "Extensible Resource Management for Cluster Computing", In
Proc. of the Seventeenth IEEE International Conference on Distributing
Computing Systems (ICDCS'97). Baltimore, Maryland, May 1997.
Nayeem Islam, Andreas L. Prodromidis, Mark Squillante, Ajei Gopal and
Liana Fong. "Extensible Resource Scheduling for Parallel Scientific
Applications", In Proc. of the Eighth SIAM Conference on Parallel
Processing for Scientific Computing (PP'97), Minneapolis, Minnesota,
March 1997.
-
Refereed Workshop Publications
Andreas L. Prodromidis and Salvatore J. Stolfo. "Cost Complexity Pruning
of Ensemble Classifiers", KDD'00 Workshop on Distributed and Parallel
Knowledge Discovery, Boston, MA, August 2000
Andreas L. Prodromidis and Salvatore J. Stolfo, "A Comparative
Evaluation of Meta-Learning Strategies over Large and Distributed
Data Sets", Sixteenth International Conference on Machine Learning
(ICML-99), Workshop on Meta-Learning, Slovenia 1999
Andreas L. Prodromidis and Salvatore J. Stolfo. "Pruning
Meta-Classifiers in a Distributed Data Mining System", KDD'98
Workshop on Distributed Data Mining, New York, New York, August 1998
(preliminary version of the NIT'98 paper).
Salvatore Stolfo, Andreas L. Prodromidis, Shelley Tselepis, Wenke Lee,
Wei Fan and Philip Chan. "JAM: Java Agent for Meta Learning Over
Distributed Databases", AAAI'97 Workshop on AI Approaches to Fraud
Detection and Risk Management, Providence, Rhode Island, July 1997
(preliminary version of the KDD'97 paper).
Salvatore J. Stolfo, Wei D. Fan, Wenke Lee, Andreas L. Prodromidis
and Philip K. Chan. "Credit Card Fraud Detection Using Meta-Learning:
Issues and Initial Results", AAAI'97 Workshop on AI Approaches to
Fraud Detection and Risk Management, Providence, Phode Island,
July 1997
Nayeem Islam, Andreas L. Prodromidis and Mark S. Squillante,
"Dynamic Partitioning in Different Distributed Memory Environments",
In Proc. of the IEEE IPPS/SPDP'96 Second Workshop on Job Scheduling
Strategies for Parallel Processing. Hawaii, April 1996.
-
Other Publications
Andreas L. Prodromidis, Salvatore J. Stolfo, Shelley Tselepis,
Terrance Truta, Jeffrey Sherwin and David Kalina, "Distributed Data
Mining: The JAM system architecture", Technical Report, CUCS-007-01,
Columbia University, 2001
Andreas L. Prodromidis and Salvatore J. Stolfo, "Agent-based
Distributed Learning Applied to Fraud Detection", CUCS-014-99
Andreas L. Prodromidis and Salvatore J. Stolfo, "Agent-based
Distributed Learning Applied to Fraud Detection", CUCS-014-99
Andreas L. Prodromidis, "On The Management of Distributed Learning
Agents", Technical Report, Columbia University, Computer Science
Department, CUCS-032-97, 1997
Sheikh Hidayatullah, Nayeem Islam, Andreas L. Prodromidis, Vineet
Singh, "A Mechanism for Combining Data Analysis Algorithms with
Databases on the Internet", Technical Report, IBM, RC 20713 (1996)
Nayeem Islam, Andreas L. Prodromidis, Mark S. Squillante, Liana Fong
and Ajei Gopal, "Octopus: Extensible Resource Management for Cluster
Computing", Technical report IBM, RC 20526, 1996
Nayeem Islam and Andreas Prodromidis, "A design pattern for adaptive
systems", Technical Report, Columbia University, Computer Science
Department, CUCS-039-95, 1995.
Andreas L. Prodromidis, Panagiotis Heliotis and Sofia Kontaxaki,
``SHAPE: The Complete Reference'', Technical Report, Intrasoft Corp,
Athens, 1993
-
Papers in Progress
Andreas L. Prodromidis, Salvatore J. Stolfo and Philip K. Chan,
"Effective and Efficient Pruning of Meta-Classifiers in a Distributed
Data Mining System"
-
INVITED PANELS, PRESENTATIONS AND POSTERS
- Invited Speaker, "JAM: Java Agents for Meta-Learning for
Fraud Detection", CASCON'97 conference, Toronto, Canada, November 1997
- Invited Panelist, Secure E-Payments, CASCON'97 conference.
Toronto, Canada, November 1997
- Demonstration of the JAM system , CASCON'97 conference,
Toronto, Canada, November 1997
- Invited Poster, "The JAM project", CASCON'96 conference,
Toronto, Canada, November 1996
-
SOFTWARE RELEASES
- JAM version 2.11 , Columbia University, June 1998
Latest release to DARPA and the Public Domain.
Earlier releases, include versions JAM-1.0, JAM-1.1, JAM-2.0 and
JAM-2.1
-
PATENTS
"Method for obscuring access patterns using a buffer memory", J. Smith, J. Chung, A. Prodromidis, J. Sherwin, S. Stolfo, iPrivacy invention disclosure, July 2000
"Flexible Dynamic Partitioning of resources in a cluster Computing Environment", L. Fong, N. Islam, A. Gopal, A. Prodromidis, M. Squillante, IBM invention disclosure, May 1997
"Gang Scheduling for Resource Allocation in a Cluster Computing Environment", L. Fong, N. Islam, A. Gopal, A. Prodromidis, M. Squillante, IBM invention disclosure, November 1997
-
COMPUTER SKILLS
- Programming Languages
C, C++, Java, Perl, Basic, Pascal, Prolog, Lisp, Fortran. Visual C/C++ conversant.
- Operating Systems
UNIX (Solaris, AIX, BSD), Linux, Windows 2000/NT/98/95, DOS.
- Other Technical Skills
SQL, Stored Procedures, Oracle8i, MySQL.
Apache Servers, Apache Modules, CGI, XML, HTML, Perl DBI, CORBA.
Java RMI, Multi-thread programming, JDBC, Java Applets, AWT, WBI.
System Administration, qmail, sqwebmail, Firewall Administration.
Unix shell scripting, awk, InstallShield.
-
PROFESSIONAL SOCIETIES
AAAI, IEEE Computer Society, ACM, Engineering Chamber of Greece
-
ACTIVITIES AND SERVICES
- Program Committee Member PKDD'01 Workshop on Mobile and Distributed Data Mining
- Program Committee Member ECML'00 workshop on
Meta-learning
- Program Committee Member ICML'99 workshop on
Meta-learning
- Reviewer for journals, books, conferences, workshops
Journal on Distributed and Parallel Databases
Journal on Knowledge Discovery and Data Mining
ML Journal special issue on Integrating Multiple Learned Models
Object Oriented Application Frameworks Wiley's book, 1998
KDD'98, ICML'98
IEEE Computer Theme Issue on
Recent Developments on Operating Systems
CACM Theme Issue on
Managing Object-Oriented Software Development Projects
CACM Theme Issue on Design Patterns
AAAI and KDD Workshops
- Ph.D. Committee Member , Computer Science Department,
1995-1997
-
REFERENCES
Available upon request