A cloud based deployment project that was implemented using Microsoft Azure cloud service. This project is mainly focused about customer segmentation process using unsupervised machine learning algorithm called as K-means Clustering. I have developed this project in google colaboratory notebook and approached building the algorithm step by step as starting from Exploratory Data Analysis(EDA) to K-means algorithm with featuring of data visualization step to show the clusters around the centroid. As this is my first cloud-based project , after completing the algorithm , I have then moved to the cloud system called as Azure for using machine learning service to create an model and deploy it as the real time endpoint. I created a machine learning workspace for initial registration of my model with specifying resource group under the estimated location for configuration.
Azure is a vast cloud system that was developed by Microsoft by including many features like storage , networking , databases , AI & Machine Learning , Analytics , DevOps and many more to implement at large scale integration and development of the rapid growing technologies.
Above is the list of the services that can be implemented for any projects at greater time of initiation. As you cann see Machine Learning from Analytics and IoT core
It is the process of organizing customers into specific groups based on shared characteristics, behaviors, or preferences, with the aim of delivering more relevant experiences.
K-means Clustering algorithm used here in the project
Programming language used in the notebook : Python
Services used from Azure
- Compute Instances -> Virtual Machine
- AI & Machine Learning(Analytics) -> Machine Learning
List of steps undergone in this project
Creating and configuringas following
-
Computing instances.
-
Configuring and deploying a virtual machine that is created.
-
Machine Learning workspace.
-
Directing to machine learning studio as the web version available.
-
Importing the model and performing testing & training in machine learning studio environment.
-
Creating the endpoints and setting up the deployment specifications
-
Finally deploying the machine learning model as the real time endpoint
-
Model has been deployed in the Azure ML environment
Below are the sample screenshots of the services used in this deployment
Azure cloud console
Azure Machine Learning Studio
Kamalesh Selvaraj , Shreya Bodla
The complete procedure of this work has been uploaded in the form of pdf soon into this repositery as this project requires a final development with renamed as Machine Learning model - Deployment in Azure