Data visualization and analytics
In this phase, you can continue the data exploration through various analytics and visualization tools to assess the data fitment for ML training post profiling. You can continue to leverage services such as Amazon Athena, Amazon Quicksight, and others introduced to you in Chapter 6, Processing and Consuming Data on the Cloud.
Feature engineering (FE)
In this phase, your responsibilities as IoT professionals are very limited. This is where the data scientists will determine the unique attributes in the dataset that can be useful in training the ML model. You can think of rows as observations and columns as properties (or attributes). As data scientists, your goal is to identify the columns that matter in solving a specific business problem (aka features). For example, with image classification, the color or brand of a car is not a key feature to determine it as a vehicle. This process of selecting and transforming variables to ensure the creation...
