ML 1.5Lesson outlineLesson 5 of 5 in Module 1
Where Data Comes From
See how collection and labelling decide what a model can learn, before any algorithm runs.
Free sample. Everything in this module opens with a free account.Video in production.
Lesson outline: Where Data Comes From
In this lesson9 parts
- 01What the messages taught the rule
- 02How the five thousand five hundred and seventy four messages were collected
- 03Reading the collection as all text messages
- 04Collection traces in the spam and the ham
- 05Duplicates, and the other data-quality checks
- 06Labels made by people
- 07Training data against the data a model serves
- 08The code
- 09Recap, your challenge, and what comes next
Builds on
This lesson stands on its own. These go deeper into what it uses.
Key terms: Where Data Comes From
The words this lesson introduces, each in one line. The module’s glossary collects them all, and the AI glossary has every term across the courses.
- human labelling
- People deciding the label of each example.
- sampling frame (reminder, Stats 1.7)
- The set of units that could end up in a sample; here, four particular collections of messages.
- selection bias (reminder, Stats 1.7)
- When the way rows are chosen makes the data differ from the population a claim is about.
- survivorship bias (reminder, Stats 2.3)
- A summary built only from the cases that passed a selection step: the spam people complained about.
- collection artifact
- A pattern in the data made by how it was collected, not by what it records.
- duplicate
- A row whose content repeats another row.
- data-quality check
- A test of the data for problems, run before any model.
- annotator
- A person who labels examples.
- label noise
- Labels that are wrong.
- train/serve mismatch
- A difference between the data a model was trained on and the data it runs on.
Quiz: Where Data Comes From
Your first pick on each question is the one that counts. Getting one wrong here is how the lesson sticks.