Giter VIP home page Giter VIP logo

diva-2's Introduction

  1. Generate many synthetic datasets -> save these locally in a folder somewhere
  2. Split data into test/train sets -> save those seperately
  3. Poison/add noise to the training sets -> train model on clean data, save its accuracy. Train model on poisoned data, save its accuracy. (can save clean + posisoned model accuracy together if wanted). Generated test datasets should be able to be posioned by any number/combination of poisoners.
  4. Compute the complexity measures of the poisoned datasets -> save that
  5. Create a dataframe where complexity measures (as a 1D array) map to its clean accuracy score: c_measure -> accuracy_clean. This should be a large dataframe where each entry is (c_measure -> accuracy_clean). Save this as the meta database
  6. Train a meta-learner on the meta-database. Whatever classifier is alright (just use the one DIVA uses).
  7. We have a meta-learner, now test out DIVA.
  8. Generate more synthetic datasets
  9. Poison them
  10. Compute their C-measures
  11. Feed C-measures into the meta-learner which will spit out an estimated accuracy that our classifier should reach (if it were non-poisoned)
  12. Actually train a classifier on the poisoned data and save its accuracy
  13. Compare its accuracy to the meta-learner's estimated accuracy, if large it is poisoned.

diva-2's People

Contributors

tinchuu avatar luorixo avatar

Watchers

avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.