Skip to content

Modeling Some Metrics of Advertisement's Data with Probability Distributions

Notifications You must be signed in to change notification settings

IzkSensei/Data-Modeling

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Modeling Metrics with Probability Distribution

Probability distribution can simulate various series of events. Thus, it can be used in order to simulate some events, behavior of the system, or even predict some actions. Different distributions can be used for different usage. In this project, some queries were calculated by some probabilities.

Dataset

This dataset consists of the actions of users on an online market including:

  • Action: Click/load
  • Time
  • Event ID
  • Device ID
  • Page Offset
  • Post List
  • Post ID

    Problems & Solutions

    1- Probability of clicking on the ad?

    The clicking on ads can be modeled by Bernoulli distribution where the user can click or not.

    2- The rank of the first ad that is clicked?

    The rank of the first ad that is clicked on it can be simulated with the geometric distribution where we try until the first success.

    3- Average distance between clicks?

    For calculating the average distance between clicks of users, exponential distribution can perform well. It is used for modeling distances between events like distances between clicks.

    4- The probability of clicking on the first 3 ads?

    In order to calculate the probability of clicking on the first 3 results, a commutative geometric distribution is suitable.

    Data analytics

    Some analytics is done on the dataset to know better insights from the dataset in the below. In addition, the Sankey diagram was used in order to demonstrate dataset characteristics.

    The Sankey diagram with Plotly:

    The HTML file is here.

About

Modeling Some Metrics of Advertisement's Data with Probability Distributions

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published