PX-02
PrototypeMachine-learning experiment
Weather Markets.
A machine-learning experiment for Kalshi's daily weather markets.
Record / In development
- Scope
- Daily weather
- Method
- Model-driven
- State
- Active prototype
System path / 04 stages
01
Question
Define a narrow market hypothesis
02
Model
Test a weather-driven signal
03
Market
Compare the signal with available prices
04
Review
Measure outcomes and revise
Decision notes / 03
D-01
Keep the market narrow
The experiment is limited to daily weather contracts so model behavior and outcomes remain inspectable.
D-02
Treat it as research
The project is presented as an active prototype; implementation notes and results will be added as they stabilize.
D-03
Separate signal from outcome
The useful question is whether a model adds information, not whether an individual market resolves favorably.