How To Deliver Data Science Projects ?
Machine Learning vs Traditional Modelling/Programming
Data science projects deliver services that predict target variables to be used later by end-users or business stakeholders. The input data can be either situational or temporal, and the prediction service can predict the targets individually (one instance per time), in groups, or in big batches. The predicted targets can be delivered as instant responses using live requests or delivered as synchronised data batches that get updated according to a predefined schedule. This article discusses the various ways to deliver an AI/ML project.
Machine learning models are typical functions that follow standard API practices, except they have special operational requirements due to their experimental nature and due to the continuously changing nature of data in the runtime.
Machine Learning Delivery Methods:
The delivery method for data science projects vary according to the problem’s nature and the ecosystem infrastructure.
- Restful API can be called as needed by the consumer, who passes the input parameters and receives the output value. This can be used for integration with the front end for live recommendation or live response.
- Batch API that…
