Professor at Unicamp
Brazil
- Current
-
- Professor at UNICAMP
- Consultant - R&D project at Elektro
- Consultant - R&D project at AES - Tietê
- Past
-
- Research associate at State University of Campinas, Department of System Engineering
- Researcher and programmer as part of a R&D project at CPFL
- Education
-
- Universidade Estadual de Campinas
- Universidade Estadual de Campinas
- Universidad Nacional de Ingeniería
- Connections
-
33
connections
- Industry
- Research
- Websites
Ivette Luna’s Summary
Ivette Luna was born in 1978, in Lima, Peru. She received her B.Sc. degree in Mechatronic Engineering from the National University of Engineering (UNI), Peru, in 2000, and her M.Sc. and PhD degree in Electrical Engineering from the State University of Campinas (UNICAMP), Brazil, in 2003 and 2007, respectively. Among her main topics of interest are time series analysis and forecasting, neural networks, fuzzy inference systems and learning techniques.
Ivette Luna’s Specialties:
Development of time series models. Modelling strategies based on neural networks, fuzzy systems and evolving rule-based systems. Data driven approaches for automatic selection of time series models in a large scale.
Ivette Luna’s Experience
-
Professor
UNICAMP
(Non-Profit; Electrical/Electronic Manufacturing industry)
August 2009 — Present (6 months)
Research fellow at the Institute of Economics. Lecturing Mathematic for economists II.
-
Consultant - R&D project
Elektro
(Research industry)
May 2008 — Present (1 year 9 months)
Time series models for energy demand forecasting using Computational Intelligence tools (neural networks and fuzzy systems) and BJ models, envolving different macro-economic variables for a multi-step ahead forecasting task.
-
Consultant - R&D project
AES - Tietê
(Education Management industry)
February 2008 — Present (2 years )
Development of daily and monthly inflow models for multi-step ahead forecasting using fuzzy inference systems and java programming.
-
Research associate
State University of Campinas, Department of System Engineering
(Educational Institution; 10,001 or more employees; Higher Education industry)
November 2007 — August 2009 (1 year 10 months)
Developing time series forecasting systems using computational intelligence techniques. Java programming. Research.
-
Researcher and programmer as part of a R&D project
CPFL
(Privately Held; 1001-5000 employees; Utilities industry)
June 2005 — December 2006 (1 year 7 months)
The R&D (Research and Development) project was entitled "A new methodological approach for studying short, mid and long term energy demand forecasting".
As part of the "mid term" team, I developed a prediction system based on neural networks and mixture of experts models, adjusted for calculating the future twelve monthly energy demand values for the company, applying a "down-top" strategy. This strategy consisted in forecasting the energy demand for each region covered by the company, as well as for each class of consume, that agregated, gave as a final result the total energy demand for the next year.
The project also provided a better understanding of the energy market as well as its realtionship withh other explanatory variables (economic, populational, etc).
Ivette Luna’s Education
-
Universidade Estadual de Campinas
Phd , Electrical Engineering , 2003 — 2008
Dissertation titled "Time series analysis and modeling based on fuzzy rules".
Scholarship given by the State of Sao Paulo Research - Fapesp.- Activities and Societies:
- The thesis proposes a novel adaptive/evolving fuzzy inference system, used for multi-step ahead forecasting of different time series, including weekly and monthly riverflood time series, financial time series and other benchmark time series used in the literature.
-
Universidade Estadual de Campinas
Msc , Electrical Engineering , 2001 — 2003
Dissertation titled "Recurrent neurofuzzy networks applied for systems identification and control".
Scholarship given by the The National Council for Scientific and Technological Development - CNPq.- Activities and Societies:
- The dissertation proposes different recurrent and non-recurrent hybrid models for system identification and control, as well as a rich comparison with other proposals of the literature, using different neural and fuzzy architectures for a set of benchmark problems.
-
Universidad Nacional de Ingeniería
Graduate , Mechatronics , 1996 — 2000
A year scholarship for studying english.
Active student for organizing annual science expositions.
Engineering award for best robotic project at CONEIMERA 2000.
Additional Information
Ivette Luna’s Websites:
Ivette Luna’s Interests:
neural networks, evolving fuzzy models, time series forecasting (financial time series, streamflow time series, energy demand time series), professional networking. Reading, culinary, dancing, swimming, travels.
Ivette Luna’s Groups:
International Neural Network society (INNS).
International Institute of Forecasters (IIF).
-
Institute of Business Forecasting & Planning -
Machine Learning Connection -
Universidad Nacional de Ingenieria (UNI)
Ivette Luna’s Honors:
2008 5th place at the International Neural forecasting competition (NN5) - complete dataset (111 time series forecasted using a data driven automatic strategy) - http://www.neural-forecasting-competition.com/ presented at the International Symposium of Forecasting - ISF'08, Nice - France.
2007 8th place at the International Neural forecasting competition (NN3) - reduced dataset - http://www.neural-forecasting-competition.com/NN3/index.htm presented at the International Joint Conference on Neural Networks - IJCNN'07, Orlando FL United States.
2000 Fist place, best automation project. VII Congreso de Ingenieria Mecanica, Electrica, Electronica y Ramas Afines, CONEIMERA'00, Cusco - Peru.
Ivette Luna’s Contact Settings
Interested In:
- career opportunities
- consulting offers
- new ventures
- job inquiries
- expertise requests
- reference requests
- getting back in touch
