Operations Research and Management Science (ORMS) form an interdisciplinary branch of applied mathematics, economics, engineering and the sciences. ORMS professionals develop and use scientific principles, strategies, and methods to improve an organization’s ability to implement high-quality operational and management decisions.
Within this broad context, our company provides advanced mathematical modeling and optimization services from ’A’ to ’Z’ to its worldwide client communinity.
The company has been founded by Dr. E. Bajalinov in Hungary, 1995. We work with leading edge developer partners, including PhD level individual experts and use the state-of-the-art theory, methods, software and techniques.
Our products assist our clients to analyze and to solve a significant range of decision problems arising in engineering, economics, finance, and the sciences. In cooperation with our research and developer partners, we also offer more general model development, optimization and statistical services in the broad area of ORMS.
Our primary expertise is related to linear- and non-linear systems analysis and optimization (especially optimization models with fractional objectives which allow optimize such important efficiency-like criteria as return-rate) This specifically includes the development of algorithms, models, and software for global and local optimization, with extensions towards handling also a mixture of continuous and integer decision variables.
Also our main expertise includes statistical analysis for seasonal and non-seasonal time series from the real-life applications. Along with such classical and conventional methods as exponential smoothing, we use the own techniques and procedures (Walsh-transformation, optimal linear and polynomial smoothing, etc.) to predict future values and determining forecast for required confidence intervals.
Our methods have been proven and checked using numerous real-life hardest test data collections, such as
and result highly accurate predictions often over performing conventional methods and procedures.
For example, our numerical experiments performed in R-Studio on test data collections M1 and M3 (for 240 monthly time-series) stably demonstrate more accurate predictions for long-range forecasting horizon. Compared to the “conventional”, which stands for the results obtained from “the-state-of-the-art” generic function forecast() (version 7.1) in RStudio (version 0.99.903).
For technical details and further information related to Dr. E. Bajalinov’s work (with co-authors as appropriate), please check out the following link: http://zeus.nye.hu/~bajalinov/
Thanks for your interest.
Dr. Erik Bajalinov