Building Machine learning model as API

This repo has sample code to create an API to predict if a domian is DGA or not using AWS sagemaker and apigateway. It also have terraform script for deployment.

Data Setup

DGA generations code

I have taken old code from these two repo and changed it to python 3 version https://github.com/andrewaeva/DGA/blob/master/dga_wordlists/main.py

https://github.com/baderj/domain_generation_algorithms

New code and docker file to generate DGA are in: https://github.com/aftabalam01/machinelearningpipeline/tree/master/src/dataset

NoteBook to clean up and create final data

https://github.com/aftabalam01/machinelearningpipeline/blob/master/src/notebooks/Benign_dataSetup_explore.ipynb

Build and train model

https://github.com/aftabalam01/machinelearningpipeline/blob/master/src/notebooks/Build-deploy-dga-xgboost_exp5.ipynb

Sample code with tensorflow and using sagemaker script mode https://github.com/aftabalam01/machinelearningpipeline/blob/master/src/notebooks/Build%20Tensorflow%20LSTM%20model-sagemaker.ipynb

model script https://github.com/aftabalam01/machinelearningpipeline/blob/master/src/notebooks/tf-model.py

Lambda functions to integration between apigateway and model endpoint

https://github.com/aftabalam01/machinelearningpipeline/tree/master/src/lambda

Finally test code to test endpoint

https://github.com/aftabalam01/machinelearningpipeline/blob/master/src/Test/TestCode.py