Unibg International

Department of Management, Economics and Quantitative Methods

Stochastic Programming School 2007

Area bacheca: 636&


Department of
Mathematics, Statistics, Computing and Applications

10 - 20 April 2007
Università degli studi di Bergamo
e-mail: spschool2007@unibg.it
==> e-learning

Project Topic and Objectives Scientific Originality School Program Young Researchers Contributions School Schedule
Organisation and Management Sponsorships Location and Infrastructure Registration European Added Value

Proceedings Rules

1. Project Topic and Objectives

We propose an advanced programme from April 10 (Tuesday, after Easter) to April 20, 2007 (Friday), based on 10 half days of lecturing completed every afternoon by a set of presentations by young researchers and practitioners. The initiative is oriented to researchers, doctoral students and practitioners with a general aim to attract a significant audience to a key and rapidly growing area of mathematical programming.

The school aims also at establishing a qualified venue to enhance and promote the understanding by young scientists of the potentials of applied stochastic optimisation in areas such as finance, production planning, energy, telecommunications and clarify to leading practitioners the current state of the art in the development of stochastic optimisation techniques.

The proposal comes at a point in which the potentials of stochastic programming techniques in applied decision theory are becoming fully recognised in the industry, and the demand for advanced education programmes in this area is growing.

2. Scientific Originality

The conjunction of operations research methods, stochastic modelling and probability theory has been primarily associated with the seminal contributions of George Dantzig (1955, 1959) in the area of decisions theory under uncertainty leading to the first models of stochastic optimization in problems of military and defence. Since then a remarkable scientific effort in the fields of mathematical programming techniques, stochastic processes theory and computer science has facilitated on one side the adoption of increasingly powerful solution techniques to high dimensional decision problems and on the other side the accurate inclusion of alternative models of uncertainty in a wide range of dynamic management problems. The development is well documented and summarised in the web site www.stoprog.org, the fundamental scientific link of the community. The growing success of the Stochastic Programming symposia held every three years, alternatively in the US and Europe – last in Tucson Arizona in 2004 and next in Vienna, August 2007 –, is one of the motivations of our initiative, as further explained in the sequel.

The growing spectrum of application areas is now pushing for further developments and the solution of previously unforeseeable theoretical and applied problems.

Specifically in the areas of
  • Finance, related to the ongoing liberalisation process and the relationship between economic growth and demand for new financial instruments (pension funds, insurance, asset-liability management);
  • Energy and environment, also related to the liberalisation of the electricity market and the growing concern of the environmental consequences of the generally adopted economic development system;
  • Telecommunications, application of advance network technologies with complex input-output structures;
  • Production planning in the industrial sector, also growingly dependent on widely recognised as random inputs and outputs;
and several others.

A common feature of the above problem classes is their dependence on possibly complex stochastic dynamics and the inclusion of different types of risk-reward tradeoffs in the objective functions, easily leading to large-scale decision problems.

More recently the advances in the theory of risk, leading to the establishment of an axiomatic theory of risk measures, has further enhanced the understanding of the implications of alternative risk measures in decision planning problems under uncertainty.

The explicit inclusion of several thousand possible economic scenarios upon which an optimal decision strategy will depend calls for the adoption of high performance computing in the generation and solution of the associated optimization problem.

Finally the inclusion of chance constraints and integer variables may often required the formulation of the problem as a large scale mixed integer problem, whose solution will require dedicated numerical techniques.

The school is primarily proposed in order to attract and form a large number of promising young researchers and practitioners in the area of stochastic optimization and also give them the opportunity to present their research in a highly qualified venue. The two weeks training has been structured in order to facilitate the learning process from a theoretical and applied point of view and help an effective knowledge transfer between scholars in the field, young researchers and practitioners.

3. The Spring School Program

We propose two intense weeks of training plus a sequence of afternoon sections devoted to presentations by school attendants.

The programme combines theory and applications in finance, energy, telecommunications, production planning and other application areas where dedicated developments did recently took place. Theoretical lectures are essentially proposed in the morning and the afternoons are left to lectures strictly related to different application fields and young researchers presentations.

As further clarified in section 3.1, the school has been structured in order to cover topics in the following streams:

  • Stochastic optimization, introduction and key results, primal and dual theory
  • Scenario generation and approximation of stochastic processes
  • Integer, mixed integer, chance constrained, non standard linear and non linear stochastic programming
  • Stochastic control and dynamic stochastic programming
  • Model generation and solution methods for different problem classes

Theoretical sessions are finally linked to the different application areas by a number of lectures specifically addressing the numerical implications of different modelling approaches.

To enhance the attendant’s interests and promote a discussion on several key aspects of stochastic programming applications, we propose a number of specific lectures in which the future of stochastic programming developments is discussed in a critical way, and possible ways forward presented from different perspectives. In this specific context prof. Vladimirou will present his view on future SP developments in finance, prof. Tomasgard on system developments and user requirements, prof. Dempster on application fields and model developments, prof. Dupacová on non-standard SP techniques, prof. Gaivoronsky on software developments from different application areas. In order to facilitate the interaction with the students, a group of local (from the organising committee) and foreign scholars will be present beyond the required teaching days. Prof. Pflug, main organiser of the Stochastic Programming symposium to be held in Vienna in August 2007 (www.spxi.org) is cooperating in order to maximise the synergies with the proposed school.

• List of Speakers

3.1 Young Researchers Contributions

The School will leave space every afternoon to theoretical and applied contributions (three per day, for a total of 27 talks over the entire program period) by young researchers and practitioners attending the school in the general domain of Decision making under uncertainty related to:

  • Mathematical programming and dynamic optimisation in particular
  • Two-stage and multistage stochastic programming
  • Stochastic control
  • Dynamic risk measures
  • Model generation and software developments
  • Solution techniques for linear and nonlinear programmes
  • Sampling techniques for large scale problems
  • Output analysis for large-scale decision problems.

Submitted papers, in the spirit of the school, can be theoretical or, preferably, addressing decision and valuation problems in application areas such as economics and finance, energy, telecommunications, industry, system biology etc.

Unpublished articles or extended abstracts must be submitted before January 31, 2007 to spschool2007@unibg.it following the scheme in the web page. Acceptance will be given shortly after the deadline.

Accepted papers will be included in the programme by subject, following the school evolution.

The accepted contributions will be considered for publication in a special issue of the IMA Journal of Management Mathematics (Oxford University Press). Talks are expected to last 40 minutes to leave 5 minutes for comment and discussion.

3.2 The School Schedule

The two-week course is based daily on four extended 60-minute lectures on "Theory and Applications" of stochastic programming to be held at 9:00 am, 10:00 am, 11:30 am and 2:00 pm, and three 40-minute talks starting at 3:15 pm (see the program at a glance).

• Detailed Schedule

The schedule at a glance and its time structure:

Week 1

Week 2

4. Organisation and Management

The Organising Committee is based in Bergamo (Italy) and formed by
Giorgio Consigli (Main organiser, University of Bergamo)
Marida Bertocchi (University of Bergamo, IT)
Vittorio Moriggia (University of Bergamo, IT)
Elisabetta Allevi (University of Brescia, IT).
The programme has been structured in cooperation with the Scientific Committee:
Elio Canestrelli (University Ca’ Foscari of Venice, IT)
Michael A.H. Dempster (University of Cambridge, UK)
Jitka Dupacová (Charles University of Prague, CZR)
Georg Pflug (University of Vienna, AU).

Pioneer Investments, one of the largest Fund managers worldwide with a definite and strong applied research vocation, will support the SP school through the accommodation sponsorships offered by the Organising committee.
It has been agreed that during the two week programme Pioneer representatives will be given the opportunity to present the Group activity (primarily based in Milan, Dublin and Boston) to the school attendants and interview selected young researchers and doctoral students attending the school, to be considered, in full coordination with the respective University supervisors and academic coordinators, for inclusion in their high potential human resource database.
School attendants with an interest in a finance career are required to send their CV's, including for knowledge their academic supervisors, to spschool2007@unibg.it expressing an interest to be interviewed.

We are also glad to inform that OptiRisk Systems (www.optirisk-systems.com), as part of CARISMA participation to the school, has agreed to offer a 6 month license of SAMPL/SPInE to the school attendants. During the school those who intend to take advantage of this opportunity will have to fulfil an application form available at the organization desk. OptiRisk Systems offers products and services (www.optirisk-systems.com/products) in the area of Optimisation, Risk Modelling, Portfolio Planning, Asset and Liability Management, Supply Chain Management, Strategic & Tactical Management, Scheduling of Transport. We acknowledge and thank the UK-based company run by Professor Mitra and his colleagues at Brunel University, to give this important opportunity to SPS2007 attendants.

5. Sponsorships

Our school is underwritten by Scientific Contributors and Corporate Affiliates. We're grateful to all of our sponsors for their generous support.
For information on sponsorship opportunities, please email us for further information.

University of Bergamo

Department of
Mathematics, Statistics, Computing and Applications
Lorenzo Mascheroni

The Italian Association of
Mathematics Applied to Economic and Social Sciences

Pro Universitate Bergomensi


The school is endorsed by

The Stochastic Programming Community

Italian Federation of Applied Mathematics


6. Location and Infrastructure

The Faculty of Economics of the University of Bergamo – www.unibg.it/economia – provides an ideal environment for the school. A modern building realised roughly 10 years ago with fully computerised classrooms and a conference room. Depending on the attendance we can easily accommodate from 50 to 150 students during the proposed period. Lunch breaks will be organised by the faculty mensa in the same location.

A number of hotel and B&B accommodations are also available in the neighbourhood of the school location at agreed rate (for all details and required registrations see the Accommodation List).

• Trains and Buses from the Airports to Bergamo.

7. Registration

Theoretical lectures have been structured over the two weeks covering initially theoretical and modelling subjects and finally numerical and solution techniques. Applied sections alternate starting with financial applications and ending the program with industrial applications.

The attendance to the whole program is suggested and did inspire the program set-up.

It is possible nevertheless to register and attend the school for half program (first or second week): a specific interest can motivate the choice.

The deadline for early registration is January 31, 2007.

full program
half program

Table 1. Registration fees

Late registrations will be accepted with a 75 € additional fee above the selected period, after January 31, 2007.

We accept the payment through bank transfer to University of Bergamo with Banca Popolare BG – Gruppo BPU as follows:

Payments from Italy - ABI: 05428, CAB: 11101, CIN: L, c/c 21048
Payments from abroad - IBAN: IT43L0542811101000000021048 - BIC (Swift) BEPOIT22020
Reason for payment: registration to the Spring School on STOCHASTIC PROGRAMMING: Theory and Applications, April 10, 2007 - April 20, 2007.

Additional details that may be required by payment institutions or researchers are the following:

Organizing Dept.:
Address: Università degli studi di Bergamo - via dei Caniana 2 - 24127 - BERGAMO
VAT: P. IVA: 01612800167
Fiscal Code: Cod. Fisc. 80004350163

All payments must be in Euro.

Registration fees will cover the school material, the lunch and evening meals (up to 10 €), the conference dinner and the Sunday excursion. A special convention with a number of restaurants in the neighbourhood of the university will cover the evening meals.

All applicants are expected to cover the costs of their journeys.

Registered Italian and foreign academics can apply to the Organising committee to have accommodation costs covered by a sponsorship (Accommodation Form). A number of 30 accommodation sponsorships will be provided by the committee: the sponsorship will cover the cost of the accommodation in a convenient and nice location near the University over the eleven nights between Monday, April 9, 2007 and Friday, April 20, 2007.

Only early registered academics to the entire school programme can apply. Academics with presentations included in the programme will be given priority in the sponsorship assignment. Applicants will be considered on a first-in basis.

The following average costs per day per person (excluding travel costs to reach Bergamo) have been estimated to attend the school and are here reported to help Italian and foreign applicants in their financial planning:

  • Accommodation: in nearby B&B’s and hotel, according to the Accommodation List, from 50 € (This list is for all applicants to the SPS2007 who did not apply for the accommodation sponsorship).
  • Evening meal: from 10 to 25 €.
  • Daily costs outside the school (transport to reach the school, drinks and breaks): 5 to 10 €.

8. European Added Value of Topic

The proposal of the School comes at a point in which the potentials of stochastic programming techniques in applied decision theory are becoming fully recognised in the industry and the demand for advanced education programmes in this area is growing. The growing spectrum of application areas is now pushing for further developments and the solution of previously unforeseeable theoretical and applied problems. Specifically in the areas of finance – related to the ongoing liberalisation process and the relationship between economic growth and demand for new financial instruments (pension funds, insurance, asset-liability management) –, energy and environment – also related to the liberalisation of the electricity market and the growing concern of the environmental consequences of the generally adopted economic development system –, production planning – in the industrial sector, also growingly dependent on widely recognised as random inputs and outputs –, and several others. For the first time, thanks to the Institute, internationally recognised scholars, students and practitioners will be given the opportunity over an extended period of time in a truly multidisciplinary venue, to deepen their knowledge and exchange views and suggestions in order to enhance the understanding and confidence on the potentials of an applied field whose benefits to effective decision making are still to be fully understood.

9. European Added Value of Participants

Stochastic programming is almost by definition a field at the conjunction of optimisation theory, stochastic processes and probability theory and decision theory under uncertainty. Participants will gain enormous insights in different application areas and at the same time be confronted with the challenges coming from the industry and the deep theoretical results of stochastic optimisation theory with implications on valuation theory, dynamic decision making and economic and financial planning.

We also aim with this initiative to strengthen and enlarge the area of young researchers interested in the subject of stochastic optimisation, that we see at the heart of scientific development from a computational and optimisation viewpoint.

Warning - .zip files: you have to save the file in a folder of your PC. You cannot open directly the file from the website.