Visualization for yearly income prediction

This project is an interactive part of a bigger Adults project, so you are very welcome to visit it on my website and try it on your own

A short description of what this app does under the hood:

• the Random Forest Classification model is used for training
• data from the user is collected and transformed into the format, required by the model
• model predicts the income class
• the graph with Shapley values (feature importance) is built based on the prediction

Here a new 'Country' file is used to let the user select among all countries in the world. The custom column 'developed' is added with true/false values based on Table A of this United Nations country classification report. This column is needed to transform the data to feed to the model.

The following data is collected from the user to make a prediction:

• Personal info
  • Age
  • Sex
  • Country
  • Ethnic group
• Family info
  • Marital Status
  • Family Belonging
• Education and professional info
  • Education level
  • Amount of working hours per week
  • Workclass
  • Occupation
• Capital operations Info
  • Capital gain
  • Capital loss