QUANTITATIVE DECISION SUPPORT METHODS LABORATORY

Brief Background and Mission

The Quantitative Decision Support Methods Laboratory (LAB-MAD) was initially conceived in 1998 as a Computer-Assisted Statistics Laboratory, providing a specific area for statistics-oriented software and hardware. In 2008, the laboratory also started to be used by the optimization area of the Department of Electrical Engineering, particularly for stochastic optimization. It is the combination of these two areas of study – statistics and optimization – that makes this laboratory unique.

LAB-MAD’s mission is to conduct innovative projects in the areas of statistical modeling and optimization, mainly investigating the combined use of these two tools, often within a context of risk and return analysis in the field of economics, with the objective of developing decision support tools.

LAB-MAD is also a space where students from both areas can exchange ideas and knowledge concerning their theses, dissertations and projects.

Area of Work

LAB-MAD conducts research based on statistical and optimization modeling, with a view to applications in the areas of energy, finance and economics. Some of the basic research currently underway includes:

• stochastic programming models with restrictions on risk;
• methods for breaking down large-scale optimization problems;
• non-linear and/or non-Gaussian statistical models;
• multivariate statistical models; and
• quantitative methods for risk and return analysis

A few examples of the specific research topics currently underway are:

• statistical models for renewable energy sources (wind, solar and hydro energy);
• optimal contracting of physical and financial renewable energy assets;
• optimal planning of the operation and expansion of electrical power systems;
• calculation of insurance and pension fund reserves;
• quality monitoring for electricity generating units;
• asset-liability management methods for real assets;
• pricing of energy contracts in incomplete markets;
• sizing and optimal operation of hybrid solar/diesel/network systems for free consumers.

Importance / Applications

The vertiginous increase in computer processing capacity and the volume of data seen in recent years has enabled a significant expansion in the practical applications of quantitative decision support methods. As such, LAB-MAD has a broad field of applications for the products of its research, especially in the energy and finance sectors. This is reflected in the projects undertaken at the laboratory.

Some recent research projects include:

• asset-liability management for the oil industry (Petrobras, 2009-2011);
• calculation of the IBNR reserve for mandatory auto insurance (Lider insurance company, 2012);
• persistence analysis of private pension products (LIMRA/Bradesco, 2012);
• risk and returns analysis for projects in the oil business (Petrobras, 2008-2009);
• generation quality indicators (IPEA, 2011-2012);
• stochastic optimization model for ALM for state pensions (Government of Angola/DGM, 2008-2009); and
• internal model for constituting reserves (Mongeral, 2007-2008).

Contact Information

Lead Professor(s):
Álvaro Veiga, Cristiano Fernandes, Alexandre Street e Delberis LimaAddress:
Rua Marquês de São Vicente 225, sala L435 – Gávea – Rio de Janeiro – RJ

Tel:
(+55 21) 3527-1225 (Prof. Álvaro Veiga)
(+55 21) 3527-1222 (LAB-MAD)
(+55 21) 3527-1205 (Secretaria do DEE)

Email:
labmad@ele.puc-rio.br

Website

lamps.ind.puc-rio.br