AutoML Streams
An AutoML framework for implementing automated machine learning on data streams architectures in production environments.
Installation
From pip
pip install -U automl-streams
or conda:
conda install automl-streams
Usage
from skmultiflow.trees import HoeffdingTree
from skmultiflow.evaluation import EvaluatePrequential
from automlstreams.streams import KafkaStream
stream = KafkaStream(topic, bootstrap_servers=broker)
stream.prepare_for_use()
ht = HoeffdingTree()
evaluator = EvaluatePrequential(show_plot=True,
pretrain_size=200,
max_samples=3000)
evaluator.evaluate(stream=stream, model=[ht], model_names=['HT'])
from skmultiflow.evaluation import EvaluatePrequential
from automlstreams.streams import KafkaStream
stream = KafkaStream(topic, bootstrap_servers=broker)
stream.prepare_for_use()
ht = HoeffdingTree()
evaluator = EvaluatePrequential(show_plot=True,
pretrain_size=200,
max_samples=3000)
evaluator.evaluate(stream=stream, model=[ht], model_names=['HT'])
More demonstrations available in the demos directory.
Development
Create and activate a virtualenv for the project:
$ virtualenv .venv
$ source .venv/bin/activate
$ source .venv/bin/activate
Install the development dependencies:
$ pip install -e .
Install the app in "development" mode:
$ python setup.py develop
Paper
https://arxiv.org/abs/2106.07317
@article{DBLP:journals/corr/abs-2106-07317,
author = {Alexandru{-}Ionut Imbrea},
title = {Automated Machine Learning Techniques for Data Streams},
journal = {CoRR},
volume = {abs/2106.07317},
year = {2021},
url = {https://arxiv.org/abs/2106.07317},
eprinttype = {arXiv},
eprint = {2106.07317},
timestamp = {Wed, 16 Jun 2021 10:42:19 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2106-07317.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
author = {Alexandru{-}Ionut Imbrea},
title = {Automated Machine Learning Techniques for Data Streams},
journal = {CoRR},
volume = {abs/2106.07317},
year = {2021},
url = {https://arxiv.org/abs/2106.07317},
eprinttype = {arXiv},
eprint = {2106.07317},
timestamp = {Wed, 16 Jun 2021 10:42:19 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2106-07317.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}