Skip to content

jmpark0118/useR2019

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

useR2019

Datathon

  • The Data

    • The Health Nutrition and Population Statistics database hosted by the World Bank Group provides various key indicators related to health issues, population dynamics and nutrition. The data have been gathered from 258 countries over the last 57 years.
  • Objectives

    • The data challenge aims at informing us about potential progress towards the resolution of some health and nutrition issues as well as descriptions of population dynamics. Generally speaking, participants can focus on any aspect of the data and should provide any analysis they find relevant. The submission will provide a conclusion based on the analysis of the Health Nutrition and Population Statistics database. External open source data can be used to support conclusion and for illustrative purpose only. It is not allowed to use external data to perform your analysis. Some possible thoughts to get started are :
      • Distribution of the world population or in a specific area
      • Evolution of malnutrition for children under 5 years old
      • Can we identify a possible link between malnutrition and child mortality?
      • Evolution regarding HIV : is there any progress regarding safe practices?
      • Can we exhibit a correlation between education and some other indicators?
      • Can we identify the differences regarding cost of treatment and therapeutic development between various geographic areas?

Datathon - Visualizing infant mortality rates

  • Abstract

    • In the past, infant mortality rates were very high, but the rates are steadily decreasing due to recent advances in the medical technology and the industrialized economic system. Nonetheless, infant mortality is still on the high trend in many of the developing countries. To find out what causes the differences in infant mortality, we focused on the distinction between each demographic dividend. Afterwards, multiple regression analyses were conducted for variables with high explanatory power. Although the data was from 1960 to 2018, we used the data from 2012 to 2016, to reflect the most recent trends. Using this data, the mortality rate of infants was predicted for the year 2017 and compared with the actual values. We also marked it on the world map to confirm the overall predictability and to make it easier to understand. Through our study, infant mortality rates can be forecasted and we hope that governments can come up with ways to further reduce it.
  • https://jmpark0118.github.io/useR2019/


Poster - Forecasting and Visualizing Churn data

  • Abstract
    • Customer churn occurs when a customer does not like the service of an existing institution or feels that the service of another institution is better. From the agency's perspective, it is more economical to retain existing customers than to attract other customers, so forecasting customer churn is more important than ever. In this study, we conducted visualization to identify the characteristics of variables of mobile carrier termination data and made some statistical models using R. Supervised learning methods such as logistic regression, random forest, SVM and DNN were used. Using data from Kaggle, 7043 customer information with 21 features each, the results of the study show the highest accuracy of 79% in SVM when predicting churn rate and the most prominent variable for high churn rate was "the use of Online Security". We also developed a web-based platform with RShiny package that visualize the patterns of customers who would churn with have high probability. It also shows the probability of churn when a new customer's information entered. Through our research, telecom organizations can be aware of customers who are likely to churn and develop marketing strategies to keep these customers engaged.

About

useR!2019 Toulouse-France July 9-12, 2019

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages