Skip to main content

Globally local variable importance

Project description

Globally-local-variable-importance-algoritm-in-python

  • One method with two branches to calculate local variable importance with global models.
    • Since the variable importance given by scikit-learn's RandomForestRegressor only shows a global “averaged” variable importance for the whole data set and it is inadequate to represent the variable importance for certain areas or periods (local data set).
    • For several months working on this issue, I finally proposed a method with two branches and realised them in python which overcoming the shortcoming of traditional measures. With this method, we did not need to exclusively build local models or geographical approaches to estimate local variable importance and furtherly explore heterogeneity of variable importance.

Hopefully, I upload the codes to get some contribution from developers all over the world. Simutaneously, we wish to obtain some suggestions.

  • If you have questions or suggestios, please connect to the only auther Tao Li, Sichuan Univercity, China with the email:lp1559345469@gmail.com.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

glvi-0.1.2.tar.gz (6.3 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page