0

I have a mobile game dataset as below and I m expected to make analysis from users first 3 Days activity and predict/classify users’ Lifetime Values (LTV) from the user-level dataset with implementing logistic regression and randm forest models in order to predict the users' LTV.

My question is here; I havent been worked mobile game data and life time value prediction before but I know how to implement logistic regression for classification models. In this data set I couldnt understand how should I choose my target(dependent) variable and independent variables to build model? Because as I know for logistic regression there should be labelled data for target variable? I guess I m missing some information about life time value prediction.Aand I also couldnt find good enough document on google for life time value prediction with logistic regression

Could anyone help me about how can I create labelled data before running logistic regreession for life time value prediction? p.s: I ll use R for this projects

here is my data;

X userID_anonymized   arrival_time    event_time               eventType                                                        event_data platform session_no
1 0     12acc8df-d4f9 11/30/20 18:01 12/26/20 3:16             level_event {duration=30.328, level=47, progress=36, play_count=36, status=1}      ios         12
2 1     12acc8df-d4f9 11/30/20 18:01 12/26/20 3:16             level_event      {duration=0, level=64, progress=37, play_count=37, status=2}      ios         12
3 2     684c9a99-aeb0 11/30/20 15:06 12/26/20 0:53             level_event  {duration=39.537, level=7, progress=31, play_count=31, status=1}      ios          7
4 3     684c9a99-aeb0 11/30/20 15:06 12/26/20 0:53             level_event      {duration=0, level=58, progress=31, play_count=31, status=2}      ios          7
5 4     be725f9b-33c5  11/30/20 3:20 11/30/20 3:21 interstitial_impression                                              {'revenue': '0.035'}      ios          1
6 5     dd5ef583-258b  11/30/20 0:07 11/30/20 3:22       banner_impression                                             {'revenue': '0.0003'}      ios         13
  wifi_state      model operating_system totalrevenue
1          0 iPhone12,8         iOS 14.2    0.8706336
2          0 iPhone12,8         iOS 14.2    0.8706336
3          1 iPhone11,8         iOS 14.3    0.1839144
4          1 iPhone11,8         iOS 14.3    0.1839144
5          1 iPhone12,1       iOS 14.0.1    0.0362000
6          0 iPhone12,1         iOS 14.2    0.4605784
Ozgur Alptekın
  • 101
  • 2
  • 3
    I don't think there's a universally-applicable "life time value" statistic, since different businesses would have different understandings of how customers create value. If your organization doesn't have a standard definition for this, then you'll need to create your own. I'd start by conducting a literature review, and choosing a definition which best fits with my organization's goals. Here's more than half-a-million papers to get you started https://scholar.google.com/scholar?hl=en&as_sdt=0%2C47&q=lifetime+value+of+a+customer&btnG= – Sycorax Sep 06 '21 at 18:39

0 Answers0