Mullpy is a machine-learning library that mainly aim to solve multi-label problems. It is classifier independent, has many ensemble capabilities (diversity methods like bagging, random subspaces, etc.) and automated results presentation (Excel, images as ROC or class-separated info, etc.). It is fully configurable. At the moment supports Neural Networks and classifiers defined in files. It is working on python3.3.
Features
- Learning
- Testing
- Validation
- Classifier independent
- Auto-configuring some learning parameters
- Automatic display of completely classifier info in excel format
- Automatic display of figures (ROCs, scatters, easy to implement new ones)
- Diversity measures for the ensembles in excel table
- Meta-Ensembles construction (real outputs, oracle outputs)
- SMV/WMV/ easy to implement new ones
- Modular code that permits quick expansion
License
GNU General Public License version 3.0 (GPLv3)Follow mullpy
Other Useful Business Software
Stigg | SaaS Monetization and Entitlements API
A monetization platform is a standalone middleware that sits between your application and your business applications, as part of the modern enterprise billing stack. Stigg unifies all the APIs and abstractions billing and platform engineers had to build and maintain in-house otherwise. Acting as your centralized source of truth, with a highly scalable and flexible entitlements management, rolling out any pricing and packaging change is now a self-service, risk-free, exercise.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of mullpy!
