Logistic churn prediction models for mobile games
| Autor: | Šarlija, Anđela |
|---|---|
| Přispěvatelé: | Martinić Bilać, Tea, Perišić, Ana, Šubašić, Aljoša |
| Jazyk: | chorvatština |
| Rok vydání: | 2023 |
| Předmět: | |
| Popis: | Logistička regresija je statistička metoda koja se koristi u analizi podataka u kojoj je cilj opisati vezu izmedu dihotomne zavisne varijable i jedne (ili više) nezavisnih varijabli. U ovome radu prikazana je primjena logističke regresije u modeliranju problema odljeva korisnika u industriji mobilnih igara. Nakon prikaza osnovne teorijske podloge, izgrađeni su univarijatni i multivarijatni logistički modeli za problem predikcije odljeva korisnika pri čemu su uspoređeni modeli s obzirom na različite definicije odljeva korisnika. Prikazana je izgradnja multivarijatnih modela temeljem stepwise procedure uz podršku statističkog programa R. Logistic regression is a statistical method used in data analysis where the goal is to describe the relationship between a dichotomous dependent variable and one or more independent variables. This paper presents the application of logistic regression in modeling churn problems in the mobile game industry. This work provides the presentation of the basic theoretical basis and the application based on the real-world data where univariate and multivariate logistic models are built for the problem of predicting user churn. Also, models are compared at different levels of churn definition. The construction of multivariate models based on the stepwise method is presented with the support of the statistical program R. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|