/automation-repo

Machine learning and process automation

Primary LanguageJupyter Notebook

New samples are posted here: https://github.com/katanaml/sample-apps

Process automation with Machine Learning.

  1. invoice-automation-d1.ipynb - date is split into multiple columns
  2. invoice-automation-d2.ipynb - instead of splitting date, using date difference in days
  3. invoice-risk-model-local.ipynb - step by step notebook to run xgboost on premise
  4. diabetes_redsamurai_db.ipynb - notebook which demonstrates how to fetch training data directly from DB, prepare train/test datasets and run training with XGBoost
  5. diabetes_redsamurai_endpoint_db.ipynb - notebook which demonstrates how to use Flask to expose XGBoost ML model
  6. convnet - cat vs dog image classification model built using Python code from book: https://www.manning.com/books/deep-learning-with-python (original source code from the book on GitHub: https://github.com/fchollet/deep-learning-with-python-notebooks)
  7. forecast - future price forecast for iron/steel with Prophet model. Example how to save/load Prophet model and expose Flask API
  8. regression - Keras/TensorFlow model with regression implementation to predict report execution time, before report request is submitted
  9. tfjs-sentiment - TensorFlow.js example where Python model is reused to calculate hotel review sentiment
  10. regressiontfjs - TensorFlow.js use case example with training, transfer learning to predict business report execution time
  11. oracleml - machine learning with SQL in Oracle DB
  12. tf2.0 - ML model implemented with TensorFlow 2.0 and Keras
  13. tf-serving - TensorFlow Serving example to publish Keras model through REST
  14. forecast-lstm - Simple timeseries forecast example with true future
  15. tfjs-simple - TensorFlow.js example using TensorFlow.js API for data read and processing
  16. pipeline - ML pipeline example with Keras regression model, scheduled re-training and Flask REST API
  17. unsupervised - Unsupervised ML example with autoencoder to detect anomaly (fraud) in Health Insurance Claims

Author: Andrej Baranovskij, Red Samurai Consulting (https://redsamuraiconsulting.com)

Our Machine Learning product: https://katanaml.io/