PICTURE THIS · DAY 2 OF 5
Data Makes the Model
Today is an independent reading and annotation day. The packet prepares you for tomorrow's bias test.
Topic: Training data and data bias
Objective: I can describe collect → train → test and identify how missing variety can make a model unfair or unreliable.
Demonstration of learning: I can use evidence from the packet to explain one way a dataset could fail a new user.
Your task
- Get the three-page Picture This packet from your teacher or download it below.
- Read and annotate the examples as directed.
- Complete all five reflection responses in complete thoughts.
- When you finish, reread one response and add a specific example.
Word bank
collect · training data · class · pattern · test · data bias · variety
Success criteria
- All packet sections are complete.
- Your explanation connects a missing kind of example to a model failure.
- You use at least two words from the word bank.
- Your work is readable enough for a partner to follow tomorrow.
Sentence stem: This dataset may fail when ______ because it mostly includes ______.
RAISE Playground was created by the MIT RAISE Initiative and the Personal Robots Group at the MIT Media Lab.