R OUHOLLAH R AHMANI

R.RAHMANI@UT.AC.IR
SUMMARY
Doctorate of Computer Science, specializing in Machine Learning and Cloud Computing. Former Program Manager
at Microsoft and Amazon.com. Five U.S. Patents Pending.
E DUCATION
Ph.D. in Computer Science B.S. in Computer Science B.A. in Mathematics, Minor in
Washington University in St. Louis, Washington University in St Ar abic
Missouri, USA Louis Washington University in St. Louis
Dec 2008 Missouri, USA Missouri, USA
GPA: 4.00 / 4.00 May 2005 May 2005
GPA: 3.53 / 4.00 GPA: 3.53 / 4.00
L ANGUAGES
Human Languages: English (native), Persian (native), Arabic, Japanese
Computer Languages: C#, C++, C, Java, Visual Basic, Matlab, Java Assembly, Intel 80x86 Assembly, Lisp,
Windows API, Windows Mobile API, VHDL, Ruby on Rails
W ORK H ISTORY
Univer sity of Tehr an, Assistant Professor 2013Present
Faculty of New Sciences and Technologies, University of Tehran, Tehran, Iran
o Conducted research in areas of machine learning with applications to healthcare systems.
o Taught Masters and Doctoral courses.

Amazon.com, Technical Program Manager, Software Development Engineer 20102012

Amazon.com Incorporated, Seattle, Washington
o Five patents pending in first year. Selected by Vice President for patent development initiative.
o Drove autoclassification for the Amazon Catalog. Secured $1,000,000 to build new classification platform.
o Increased team efficiency converting 3 teams to scrum from waterfall.
o Ramped up new autoclassification team in Japan, drove daily execution and launch of new service platform

Micr osoft Cor por ation, Bing.com, Program Manager 20082010

Microsoft Corporation, Redmond, Washington
o I was part of the team that first shipped Bing, when it launched in June 2009. I drove relevance measurement
for categorized search, one of the key differentiating features at launch.
o After launch, I moved to drive planning and execution for new deep search experiences including Finance
Search, University Search, Dictionary and Thesaurus pages, Recipe Search, Auto Search, and others.
o In my last year, I created and drove new initiative for Parental Reviews in Bing Entertainment.

Univer sity of Tehr an, PostDoctoral Research Fellow 2008

Control & Intelligent Processing Center Of Excellence, University of Tehran, Iran
o Conducted research with over 10 doctoral students in areas of Machine Learning including medical image
segmentation of the brain, attention learning, and systems analysis.
O Taught Masters and Doctoral course, Technical Writing in English. Received excellent reviews.

Rahmani, 1 / 5
M AJOR P UBLICATIONS
Localized ContentBased Image Retr ieval 2008
Rouhollah Rahmani, Sally A. Goldman, Hui Zhang, Sharath R. Cholleti, Jason E. Fritts
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 30, No. 11
http://www.computer.org/portal/web/csdl/doi/10.1109/TPAMI.2008.112, Demo: http://www.cs.wustl.edu/accio

Multiple Instance Lear ning for Image Sear ch 2008

Rouhollah Rahmani
Doctoral Dissertation, Washington University in St. Louis
http://cid0f53aa73151eb98d.skydrive.live.com/self.aspx/Public/DissertationRahmani2008.pdf

MISSL: Multiple Instance SemiSuper vised Lear ning 2006

Rouhollah Rahmani, Sally A. Goldman
23rd International Conference on Machine Learning (25.5% acceptance rate)
http://www.cs.wustl.edu/~sg/finalmissl.pdf

Localized ContentBased Image Retr ieval 2005

Rouhollah Rahmani, Sally A. Goldman, Hui Zhang, John Krettek, Jason E. Fritts
7th ACM International Workshop on Multimedia Information Retrieval (invited paper)
http://www.cs.wustl.edu/~sg/accio/mir162ssrahmani.ps , Demo: http://www.cs.wustl.edu/accio

A Novel Semiinver se Solution Method for Elastoplastic Tor sion of Heat Tr eated Rods 2010
Majid Baniassadi, Akbar Ghazavizadeh, Rouhollah Rahmani, Karen Abrinia
Meccanica, International Journal of Theoretical and Applied Mechanics, Vol. 45, No. 3
http://www.springerlink.com/content/j708011483l4n604

Automatic Segmentation of Putamen using Geometr ic Moment Invar iants 2009

Moustafa Jabarouti Moghaddam, Rouhollah Rahmani, Hamid SoltanianZadeh
15th Iranian Conference on Biomedical Engineering
http://cid0f53aa73151eb98d.skydrive.live.com/self.aspx/Public/PutamenICBME2009.pdf

WebBased, Inter active, 2D/3D Medical Image Pr ocessing and Visualization Softwar e 2009
Seyyed Ehsan Mahmoudi, Alireza AkhondiAsl, Rouhollah Rahmani, Shahrooz FaghihRoohi, Vahid Teymouri,
Ahmad Sabouri, Hamid SoltanianZadeh
Journal of Computer Methods and Programs in Biomedicine, Vol. 98, Issue 2
http://www.cmpbjournal.com/article/S01692607(09)003022/abstract

A PolicyDr iven Method for IT Infr astr uctur e Selection in Power Distr ibution Automation System 2009
Alireza Fereidunian, Caro Lucas, Hamid Lesani, Rouhollah Rahmani, A.Wayne Wymore
International Review of Electrical Engineering Journal, Vol. 4, No. 6
http://cid0f53aa73151eb98d.skydrive.live.com/self.aspx/Public/PolicyDrivenIREE2010.doc

Local Image Repr esentations Using Pr uned Salient Points with Applications to CBIR 2006
Hui Zhang, Rouhollah Rahmani, Sharath Cholleti, Sally Goldman
14th ACM International Conference on Multimedia (17% acceptance rate)
http://www.cs.wustl.edu/~sg/ACMMultimedia06.pdf

MIWinnow: A New MultipleInstance Lear ning Algor ithm 2006

Sharath Cholleti, Sally Goldman, Rouhollah Rahmani
18th IEEE International Conference on Tools with Artificial Intelligence (35% acceptance rate)
http://www.cs.wustl.edu/~sg/MIWinnow.pdf

Rahmani, 2 / 5
R ESEARCH
Medical Imaging using Machine Lear ning, PostDoctor al Resear ch Fellow 20082009
Control & Intelligent Processing Center of Excellence, University of Tehran, Iran
o Lead team of masters and doctoral students to develop algorithms with application to medical diagnosis.
o Conducted and published research in blood cell image analysis, webbased imaging software systems, and
segmentation of the brain from MRI scans.

Multiple Instance Lear ning, Doctoral Research Assistant 20022008

Computer Science Department, Washington University, St. Louis, Missouri
o Conducted research and development in Multiple Instance learning, a division of Artificial Intelligence
o Developed algorithms with applications to image retrieval (CBIR) and pharmaceutical drug discovery
o Developed GEMDD algorithm. Results compete with the top Multiple Instance algorithms
o Developed MISSL, one of the first algorithms to combine Multiple Instance and semisupervised learning
o Developed new image representation to significantly improve retrieval accuracy in CBIR systems
o Independently designed and implemented Accio! CBIR system code base, now among the top image retrieval
systems in the world
o Coordinated with 2 advisors and 7 students

Computer Aided Diagnosis, Medical Image Sear ch, Doctoral Research Assistant 20052007
Electronic Radiology Lab, School of Medicine, Washington University, St. Louis, Missouri
o Conducted joint project between machine learning lab and Electronic Radiology Lab to develop algorithms
for computer aided diagnosis.
o Analyzed and worked to develop image representations appropriate for general diagnosis on MRI images.
o Worked on project to combine a set of experts, both human and computer, to better segment disease nodules.
o Participated in weekly joint lab journal club.

Pr oteinPr otein Binding Simulation, Doctoral Research Assistant 20062007

Computer Science Department, Washington University, St. Louis, Missouri
o Developed learning algorithm for clustering proteinprotein binding paths.
o Combined and extended work from molecular biology, and robotic path planning
o The path clusters aid biologists to determine biological similarities and visualize a large set of paths.
o Project serves as base for developing algorithms to simulate high dimensional proteinprotein binding.

Reinfor cement Lear ning, Undergraduate Researcher 20012002

Computer Science Department, Washington University, St. Louis, Missouri
o Performed research and development in Reinforcement Learning, a division of Artificial Intelligence, with
specific focus on convergent value function approximation in continuous state spaces
o Conducted applied research enabling robots to learn in complex state spaces, such as real world environment,
using feasible computational power and memory footprint
o Explored function approximation techniques for interpolation of discrete points in a continuous state space
o Developed and tested heuristic methodology for safe interpolation of data points in reinforcement learning

T ALKS, DEMONSTRATIONS
Talk & Demo: MISSL: Multiple Instance Semi Supervised Learning
o 23r d Inter national Confer ence on Machine Lear ning (Pittsbur gh, Pennsylvania) 2006

Talk & Demo: Accio! Localized Content Based Image Retrieval

o Waseda Univer sity (Tokyo, J apan) 2005
o Univer sity of Electr oCommunications (Tokyo, J apan) 2005
o National Univer sity of Singapor e (Singapor e) 2005

Rahmani, 3 / 5
 o 7th ACM SIGMM Inter national Wor kshop on Multimedia (Singapor e) 2005

P ROFESSIONAL ACTIVITIES
IEEE Tr ansactions on Patter n Analysis and Machine Intelligence 2010
Reviewer for article submitted for publication.
IEEE Tr ansactions on Knowledge and Data Engineer ing 2009
Reviewer for article submitted for publication.
J our nal of Patter n Recognition 2009
Reviewer for article submitted for publication.
J our nal of Machine Lear ning Resear ch 2005
Coreviewed 2 articles submitted for publication.
IEEE Tr ansactions on Neur al Networ ks 2005
Coreviewed article submitted for publication.

G RANT P ROPOSALS (C O AUTHOR )

Impr oving CBIR Accur acy thr ough SemiSuper vised Lear ning 2007
Principle Investigators: Sally Goldman
Proposal, National Science Foundation (recommended for funding)
KnowledgeBased Image Segmentation thr ough Machine Lear ning 2007
Principle Investigators: Sally Goldman, Jason Fritts
Proposal, National Science Foundation (recommended for funding)
Applying MultipleInstance Lear ning to ContentBased Image Retr ieval 2005
Principle Investigator: Sally Goldman
Annual Report 0329241, National Science Foundation

T EACHING E XPERIENCE
Technical Wr iting in English, School of Electr ical & Computer Engineer ing Spring, 2008
o Master and Doctoral level course
o Challenging course the goal was to teach writing excellence in English to foreign language students.
o Developed course material set and taught 3 hours/week
o Received excellent reviews from students and colleagues

Advanced Algor ithms, Depar tment of Computer Science Fall, 2006

Washington University, St. Louis, Missouri
o Graduate level course.
o Prepared and taught weekly recitation section 3 hours/week class time
o Held weekly office hours and helped run grading sessions
o Received high evaluation scores from students

Engineer ing and Scientific Computing, Depar tment of Computer Science Fall, 2007
Washington University, St. Louis, Missouri
o Conducted weekly MATLAB laboratory section 1.5 hours/week.
o Held weekly office hours and helped run grading sessions

Beginning Per sian I, Depar tment of Near Easter n Languages Fall, 2005
Washington University, St. Louis, Missouri
o Prepared and taught cultural section on Persian culture and language 1 hour/week class time
O Received high evaluation scores from students

Rahmani, 4 / 5
DESIGN P ROJECTS
Mobile Robot, Washington University, Fall 2005 Inter pr eter , Washington University, Summer 2003
oDesigned, implemented working mobile robot o Designed custom postfixed language to perform
oRobot tasks include obstacle avoidance, and mathematical operations
simple path planning towards a given goal o Designed, implemented parser and interpreter in
oImplemented on Motorola 68HC1 Handyboard Lisp
o Wrote and ran programs in custom language
Centr al Pr ocessor , Washington University, Fall 2003
oDesigned unified 32bit and 16bit instruction set Oper ating System Shell, Washington University, Fall
oDesigned dual threaded hardware program 2004
controller o Designed, wrote working commandline shell for
oIncorporated supercomputer vector processing Linux
oDesigned full hardware architecture with novel o Shell incorporated concurrent program execution,
memory registers to aid vector computing dynamic environment, and serverclient pipes
oImplemented working subset of design in VHDL Lear ning Algor ithms, Washington University, Fall 2003
oWrote assembler for instruction set o Designed, implemented machine learning systems
Compiler , Washington University, Spring 2002 o Neural Network, Decision Trees, Genetic Algorithm
o Designed, implemented working compiler for o Analyzed learning methods on sample data from
custom language using Java assembly CMU Machine Learning Repository
o Incorporated symbol table management, Inter net Pager system, Kent, Ohio, Summer 2001
callable function blocks, and scope. o Designed novel physical mobile pager system to
interface with Windows internet messaging
programs such as AIM at a base station
o Hardware designed in XiLinx, partially implemented
on bread board using TTL components
H ONORS I NTERESTS, H OBBIES
Deans Honorary Scholarship Poetry writing
I.E. & Goldie G. Millstone Scholarship Martial Arts: Karate, Jujitsu, Judo
Deans List Photography
Bausch and Lomb Award Philosophy: 11th18th century
AP Scholar with Honors

Rahmani, 5 / 5