Adaptive Real Time Machine Learning (artml) is a python library for building real time learning models. Many of the existing complex real time machine learning methods only rely on Incremental learning techniques limiting the true potential of Real time learning. Real time implies updating the models by only using the new data without pooling the old data again. This flexibility helps to update models in real time and also to build models using huge amounts of data. artml method has the flexibility to enhance the real time learning by giving all kind of flexibilities as mentioned below.
Incremental learning- Updating the model with new data in real time
Decremental learning- Updating the model by excluding unnecessary data in real time
Real time attribute addition- Adding new features to the model in real time
Real time attribute deletion- Deleting features from model in real time
Distribute processing- Power to build models using distributed data
Parallel processing- carrying out parallel processing using GPU's
For further information and use cases visit artml blog.
What is artml?
artml is an open source python library for building real time learning models. Although most of the functions look similar to sklearn, for applying artml specific syntaxes and procedures should be followed. This work is Python adoption for Prof. Saed Sayad (University of Toronto) research on Real time Machine Learning.
What can artml do for you?
artml lets you build real time learning linear models and perform real time data exploration. We also have the flexibility for real time feature selection and for parallel and distributed processing.
Easily install artml from the GitHub source:
First, clone artml using git:
git clone https://github.com/AdaptiveMachineLearning/artml.git
Then, import artml library into your python notebooks or files