Aktualności

PAZUR 30 - Marketing Science Methods oraz prognozowanie wyników żużla

15.11.2018 r.
1. Marketing Science Methods in R (Ming Shan, Robert Świderski from Kynetec)


Marketing science seeks to understand customers and market behaviors to help develop marketing solutions primarily through quantitative analytics. It blends some of its own methods with a broad arrange of general statistical approaches. So it is not a surprise that R can be a very good choice for marketing science just as R is one of the go-to choices for data science. We would like to discuss a few commonly used methods in marketing science including perceptual mapping, cluster ensemble, discrete choice modeling and MaxDiff scaling and the use of R to achieve them. Some additional applications leveraging the graphical strength of R and the power of Shiny will be briefly discussed. We will share our experience about some of the unique benefits of R for marketing science within Kynetec.


Ming Shan – Senior Director, Marketing and Data Sciences
Ming has 25+ years of combined experience in marketing/direct marketing research, financial modeling and academic research. His current roles are focused on marketing and data sciences, research methodologies and statistical modeling. Ming holds a B.S. in Computer Science and a MBA with concentration in Marketing. He also has an advanced degree from the University of Michigan. Ming has been living in Boston, U.S. since 2010. He is excited about using R everyday as he started about 15 years ago.


Robert Świderski – Senior Analyst, Marketing and Data Sciences
Robert has 7 years experience working with market research data. He started with reporting the results of panel studies for the automotive industry. Now within Kynetec Robert is using data science methods to help get better insights from custom research conducted for agriculture and animal health sector. He graduated from Sociology (UAM), Finance & Accounting (UEP) and Advanced Analytical Technics in Business (UEP).


">2. Prognozowanie wyników żużla za pomocą pakietu sport -- Dawid Kałędkowski


Rywalizacja w sporcie lub w grach online wymaga ciągłego aktualizowania jakości zawodników aby precyzyjnie dobierać właściwych przeciwników, ocenić konkurencyjność wydarzenia albo wycenić zakład. Algorytmy typu online pasują idealnie w sytuacji, w której napływ danych jest zbyt intensywny, obniżając obciążenie obliczeniowe. Na podstawie wyników biegów żużlowych zaprezentowanych zostanie kilka metod zaiplementowanych w R-owym pakiecie sport. Dowiemy się również, kto jest aktualnie najlepszym zawodnikiem na świecie i jak zmieniała się forma czołowych zawodników w czasie.


References:


1. Mark E. Glickman (1999): Parameter estimation in large dynamic paired comparison experiments. *Applied Statistics*, 48:377-394. URL http://www.glicko.net/research/glicko.pdf


2. Mark E. GLickman (2001): Dynamic paired comparison models with stochastic variances, *Journal of Applied Statistics*, 28:673-689. URL http://www.glicko.net/research/dpcmsv.pdf


3. Mark E. Glickman (1995): A Comprehensive guide to chess ratings. *American Chess Journal*, 3, pp. 59--102. http://www.glicko.net/research/acjpaper.pdf


4. Ruby C. Weng and Chih-Jen Lin (2011): A Bayesian Approximation Method for Online Ranking. *Journal of Machine Learning Research*,12:267-300. URL http://jmlr.csail.mit.edu/papers/volume12/weng11a/weng11a.pdf


5. William D. Penny and Stephen J. Roberts (1999): Dynamic Logistic Regression, Departament of Electrical and Electronic Engineering, Imperial College
15.11.2018 r.
godz. 18:00
sala: TBA
rejestracja przez meetup