https://github.com/pbiecek/xai-happiness

XAI use case based on world-happiness data
https://github.com/pbiecek/xai-happiness

happiness modelstudio xai

Last synced: about 1 year ago
JSON representation

XAI use case based on world-happiness data

• Host: GitHub
• URL: https://github.com/pbiecek/xai-happiness
• Owner: pbiecek
• Created: 2020-10-12T19:16:07.000Z (over 5 years ago)
• Default Branch: main
• Last Pushed: 2020-10-15T14:24:39.000Z (over 5 years ago)
• Last Synced: 2025-01-30T13:15:16.722Z (over 1 year ago)
• Topics: happiness, modelstudio, xai
• Language: R
• Homepage: https://pbiecek.github.io/xai-happiness/
• Size: 4.66 MB
• Stars: 1
• Watchers: 1
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# xai-happiness

XAI use case based on world-happiness data.

Find a live demo: https://pbiecek.github.io/xai-happiness/

![modelStudio](modelStudio.png)

## What factors correlate with happiness at national level?

1. Get data about happiness from 

https://www.kaggle.com/unsdsn/world-happiness

2. Train your favorite ML model that will predict the happiness score

3. Explain the model with [modelStudio](https://github.com/ModelOriented/modelStudio)

```r

library("DALEX")

library("modelStudio")

library("ranger")

# get data

happiness <- readRDS("happiness.rds")

# get model and explainer

model <- ranger(score~., data = happiness)

modelr <- explain(model, data = happiness[,-1], y = happiness$score)

# get model studio

ms <- modelStudio(modelr, happiness[c("Poland","Finland","Germany"),],

                  options = ms_options(margin_left = 220))

ms

```