time-series-classifier-flaskapp

Introduction

This is an implementation of Grape vs Non-grape time-series classifier. It's basically a flask-app which exposes an API to use the classifier. Check sample input-file(also train-test data) for the specified file-format(dataset is in time-series-dataset.zip)

Prerequisites

You need to have trained-model saved in this app-directory.(There is already saved model present in the directory, you can train new model by looking at time-series-classifier.ipynb)

Deployment

manifest.yml and environment.yml contains deployment configurations and dependencies.

To deploy application use : cf push

Testing thr API

To use this API make a POST call from rest-client with the input file. The body of the post call should contain key: file and value: should be uploaded file. URL will be the app's url only.(No specific paths)