ZEHRA CATALTEPE
__zehra@research.bell-labs.com__,
/cm/ms/departments/sia/zehra/
OBJECTIVE:
- Employment in industrial research with focus on machine learning,
pattern recognition, data mining, system identification,
signal processing or optimization.
- Development of applications in the fields above.
EXPERIENCE:
- 1998 -: Postdoctoral member of technical staff,
Statistics and Data Mining Dept,
Bell Labs Research
- 1992 - 1998: Research
assistant at Caltech
Investigated test inputs and input information in learning,
generalization, regularization, learning from hints and
examples and invariance hints.
Developed C and Matlab code for implementation of learning from hints,
and verification of other theoretical results.
- 1996-97: Teaching
assistant at Caltech
Suggested and supervised theoretical and empirical machine
learning projects.
Designed and programmed experiments on machine learning.
- 1990-91: Programmer at Halici Bilgi Islem Ltd., Turkey
Re/wrote Pascal code for a commercial accounting package.
- 1989-91: Math, Pascal, Physics Tutor
SKILLS:
- C,
Matlab, Perl, C++, Pascal, HTML, UNIX, Mathematica, Lisp, Prolog.
- Learning Theory,
Probability and Statistics, Pattern Recognition.
EDUCATION:
- 1998 :
Ph.D in Computer Science at Caltech (GPA: 3.7)
Thesis Advisor: Prof. Yaser S. Abu-Mostafa
Thesis Title: "Incorporating Input Information into Learning and
Augmented Objective Functions"
- 1994 : M.S. in Computer Science at Caltech (GPA: 3.7)
Thesis Advisor: Prof. Yaser S. Abu-Mostafa
Thesis Title: "The Scheduling Problem in Learning From Hints"
- 1991 : B.S. in Computer Engineering and Information Sciences at
Bilkent University, Turkey (Graduated first in the class) (GPA: 3.8)
Thesis Advisor: Prof. Varol Akman
Thesis Title: "Field: A Data Type for Automated Perception in Go"
PUBLICATIONS:
-
(1999) Zehra Cataltepe, Yaser S. Abu-Mostafa and Malik Magdon-Ismail,
"No Free Lunch for Early Stopping",
to appear at
Neural Computation.
abstract,
postscript (493K).
-
(1998) Zehra Cataltepe,
"Incorporating Input Information into Learning and
Augmented Objective Functions",
Ph.D. Thesis, California Institute of Techology.
abstract,
thesis available upon request.
-
(1997) Zehra Cataltepe and Malik Magdon-Ismail,
"Incorporating Test Inputs into Learning",
Proceedings of the
Advances in Neural Information Processing Systems 10 .
abstract,
postscript (194K).
-
(1994) Zehra Cataltepe and Yaser S. Abu-Mostafa,
"Estimating Learning Performance Using Hints",
Proceedings of the 1993 Connectionist Models Summer School,
M. Mozer et. al. (Eds.),
Lawrence Erlbaum Associates, Publishers,
Hillsdale, NJ. pp.380-386.
abstract,
postscript (433K).
-
(1994) Zehra Cataltepe,
"The Scheduling Problem in Learning from Hints,
M.S. Thesis, California Institute of Techology.
abstract,
postscript (675K).
(also available as
Caltech-CS-TR-94-09)
-
(1997) Eric Bax, Zehra Cataltepe and Joseph Sill,
"The Central Classifier Bound --
A New Error Bound for the Classifier Chosen by Early Stopping",
IEEE PACRIM, Victoria, B.C. 1997 conference proceedings.
postscript (322K).
(also available as
Caltech-CS-TR-97-08)
-
(1991) Zehra Kok and Varol Akman,
"Field: A Data Type for Automated Perception in Go",
Proceedings of International Symposium on Computer
and Information Sciences (ISCIS-VI), Turkey,
Elsevier Science Publishers,
pp 641-650.
abstract,
hard copy available upon request.
-
(1997) Ridvan A. Sahan, Don C. Albin, Nihal K. Sahan and Zehra Cataltepe,
"Modelling Of Transitional Air Flow System In A Channel
Using Artificial Neural Networks",
Sixth Turkish Symposium on Artificial Intelligence and
Neural Networks
( TAINN 97 ).
HONORS AND AWARDS:
- 1992: Earle C. Anthony Graduate Fellowship at Caltech
- 1987: Physics and Chemistry awards of Turkish Science Foundation (TUBITAK)
RESEARCH INTERESTS:
- Incorporating
test inputs
or input distribution information into learning.
- Learning from examples and hints.
- Regularization methods, early stopping, validation.
- Machine learning applications (speech, video, image recognition).
CLASSES TAKEN:
- Learning Systems ,
Information and Complexity, Pattern Recognition, Collective Computation,
Machine Learning, Computers and Common Sense Reasoning, Theory of Computing.
- Engineering Mathematics, Matrix Theory, Statistics, Probability,
Discrete Mathematics, Concrete Mathematics.
- Design of Asynchronous VLSI Circuits, Computer Graphics, Database Systems,
Advanced Data Structures, Computer Algorithms.
REFERENCES:
Prof. Yaser Abu-Mostafa,
Dr. Amir Atiya,
Prof. Alain Martin.