Create Machine Learning model and Predict Labels.

The first phase of the project was to setup a basic connection with Google cloud. In this phase, we create the machine learning model. It involves,

1. Collecting and organizing data from Drupal. User must be given flexibility to select data from different content types and fileds. The data sent to create model must be labelled, so we must use taxonomy based labelling of the data.
2. Sending data to Google Cloud Storage.The Google Cloud Storage JSON API is a simple, JSON-backed interface for accessing and manipulating Google Cloud Storage projects in a programmatic way. It is fully compatible with the Cloud Storage Client Libraries. For more information read the json api documentation.
3. Train the model using the data stored in Google Cloud Storage. Update the model using dynamic data from Drupal. Read Google Prediction API quick start documentation
4. Predict the label of new data.

References:

1. Github link of Google Cloud Client API for PHP.
   https://github.com/GoogleCloudPlatform/google-cloud-php/blob/v0.30.1/src...
2. Cloud Bucket functions
   https://github.com/GoogleCloudPlatform/google-cloud-php/blob/v0.30.1/src...
3. Google Machine Learning Engine - Online prediction JSON API documentation https://cloud.google.com/ml-engine/reference/rest/v1/projects/predict