Research Profile, Edward Tsang

Download Report

Transcript Research Profile, Edward Tsang

Edward Tsang Research
• Business applications of Artificial Intelligence
• Constraint satisfaction & optimization research
– A branch of combinatorial optimisation
– Applied to decision support and scheduling
• Computational finance & economics
– Computational intelligence + finance and economics
– Applied to forecasting, bargaining, wind-tunnel testing
• Enabling technology: heuristic search,
evolutionary computation, inferences
Edward Tsang (Copyright)
21 July 2015
Constraint Satisfaction & Optimization
• Core technologies for
transportation
optimization
– Guided Local Search
was used in ILOG
Solver’s vehicle routing
package, Dispatcher.
– BT: work force
scheduling problem.
– Honda: Multi-objective
optimization
Sponsors: BT, Honda Europe
Computational Finance & Economics
• Advanced computer science applied to finance
– More than using spreadsheets or computerisation of
accounting systems
• Research at Essex:
– Forecasting
– Automated Bargaining (game theory)
– Economic Wind-tunnel testing (market design)
• Affiliated Centre: Centre for Computational
Finance & Economic Agents (CCFEA)
Sponsors: Sharescope, BT, OANDA
Research Profile, Edward Tsang
Business Applications of Artificial Intelligence
Application
Technology
Finite Choices Decision
Support, e.g. Assignment,
Scheduling, Routing
Constraint Satisfaction,
Optimisation, Heuristic Search
(Guided Local Search)
Financial Forecasting
Genetic Programming
Automated Bargaining
Genetic Programming
Wind Tunnel Testing for
designing markets and
finding winning strategies
Mathematical Modelling, Machine
Learning, Experimental Design
Portfolio Optimisation
Multi-objectives Optimisation
Supplementary Information
Edward Tsang
Current Activities – Edward Tsang
Research Groups:
Constraint Satisfaction &
Optimisation
Computational Finance
Current Projects:
•
•
•
•
•
•
•
•
•
•
EDDIE for Forecasting – towards more complex trading strategies
Automated bargaining – finding Nash equilibrium strategies
Artificial markets – conditions for stylised facts
Credit cards market – designing bank strategies and Government policies
Market-based scheduling – for BT work force scheduling
Evolving middlemen strategies – for simple supply chains (BT sponsored)
Chance discovery – data mining for scarce opportunities
Port automated – vehicles scheduling
Multi-objective optimisation – for Honda’s industrial design
Portfolio optimisation by heuristic search
Affiliations:
Professor
Director 1/8/2009
Background, Edward Tsang
Education:
• BSc in Business Administration, Chinese University of Hong Kong
• MSc, PhD in Computer Science, University of Essex
Selected external positions:
•
Editorialship, including:
– IEEE Transactions on Evolutionary Computation
– Journal of Scheduling
– CONSTRAINTS
•
•
Chair, IEEE Computational Finance and Economics Technical Committee,
2004 & 2005
Co-chair, IEEE Taskforce on Portfolio Optimisation, 2006-
Consultancy:
Past employments:
Commonwealth
Secretariat
The Constraint Satisfaction Problem
Domains
(Values available)
Variables
(Decisions)
x1
x2
X
X
x3
x4
•
•
•
•
Constraints
On assignments
X
X
Constraint satisfaction is a decision problem
Task: make decisions without violating constraints
Sometimes you want the “best” solution
Main techniques: constraint propagation + heuristics
BT’s Workforce Scheduling
BT has many jobs to be
done in UK every day. It
has to schedule a large
number of teams to serve
these jobs, subject to time,
skill and other constraints.
Saving of 0.5% could
mean Millions of Pounds
per year. Guided Local
Search achieved the best
results in one of BT’s
challenge problems.
Technicians
Jobs
Foundations of
Constraint Satisfaction
• First book to define the scope
of constraint satisfaction
– Published 1993
• Arguably the most rigorous
book in constraint satisfaction
• All major concepts defined in
First Order Predicate Calculus
– Precise, unambiguous
– Little room for error