Data Science Hierarchy of Needs
From Upstream (root initiatives) to Downstream (consequent initiatives)
1. collect
1.1. instrumentation
1.4. external data
1.5. user generated content
2. move/store
2.1. reliable data flow
2.3. pipelines
3. explore/transform
3.1. cleaning
3.3. prepprocessing/preparation
4. aggregate/label
4.1. Analytics
4.2. Metrics
4.3. Segments
4.4. Aggregates
4.5. Features
4.6. Training data
5. learn/optimize
5.2. Experimentation
5.3. simpler ML algorithms