Data science gear: Python API for advanced machine learning algorithms built on top of sklearn, tensorflow and keras


Keywords
machine, learning, recurrent, convolutional, neural, network, named, entity, recognition, sentiment, analysis, deep, topic, detection, image-classification, named-entity-recognition, object-detection, sentiment-analysis, topic-detection
License
MIT
Install
pip install ds-gear==0.1.7

Documentation

Data science gear (ds-gear) is a Python API for advanced machine learning algorithms built on top of sklearn, tensorflow and keras and is distributed under the Mit license.

Description

  1. RNN_MTO_Classifier:
    Text scoring RNN based on Many-to-One architecture.
    Example applications include sentiment analysis, review scoring
  2. RNN_MTM_Classifier:
    Text classification RNN based on Many-to-Many architecture.
    Example applications include Named entity recognition, machine translation

Installation

Depenencies

  • Python (> 3.6)
  • Numpy
  • Pandas
  • sklearn
  • Matplotlib
  • Pylint
  • Doxypypy
  • Nltk (==3.5)
  • keras
  • tensorflow
  • keras-contrib (need to be installed prior to this package)

User installation

  1. keras-contrib installation (please refer to git repo @ https://github.com/keras-team/keras-contrib)
  2. package installation: pip install ds-gear

Source code

You can check the latest sources with the command:
git clone https://github.com/AI-Companion/ds-gear.git