Machine Learning: 2 Books in 1: Machine Learning for Beginners, Machine Learning Mathematics. An Introduction Guide to Understand Data Science Through the Business Application

Solution architecture – A description or a diagram of the data
pipeline that is used to score new data and generated predictions,
after the model has been built can be referred to as “solution
architecture”. This diagram can also provide data pipeline needed
to “retrain” the model based on new data.
Checkpoint decision – Before that start of the actual model
building process project must be reevaluated to determine if the
expected value can be achieved by pursuing the project. These
are also called as "Go or No-Go" decisions.
Stage III – Modeling

---

The goal of this stage is to find "optimal data features" for the machine
learning model, which is informative enough to predict the target variables
accurately and can be deployed in the production environment. The three
primary tasks that must be accomplished in this stage are: "feature
engineering, model training and the determination of the suitability of the
model for the production environment".
Deliverables to be created in this stage
Feature sets – The document containing all the features
described in the “feature sets section of the data definition
report”. It is heavily used by the programmers to write the
required code and develop features based on a description
provided by the document.
Model report – This document must contain the details of each
model that was evaluated based on a standard template report.
Checkpoint decisions – A decision regarding deployment of the
model to the production environment must be made based on the
performance of different models.
Stage IV – Deployment
The goal of this stage is to release the solution models to a lower
production-like environment such as pre-production environment and user
acceptance testing environment before eventually deploying the model in
the production environment. The primary task to be accomplished in this
stage is "operationalization of the model".

---