The JennyLab Subjectivity-IA project is an open-source initiative (test investigation) that explores how to calculate subjectivity without relying on machine learning models. Its goal is to offer a deterministic and explainable approach to decision analysis, especially in contexts where risks and needs must be evaluated.
What is Subjectivity-IA?
Subjectivity-IA seeks to determine if an action should be avoided based on two main factors:
Risk: The probability of a negative outcome occurring. Need: The urgency or importance of performing the action. The system evaluates events with different combinations of risk and need to decide whether they should be avoided.
For example:
Event: {'risk': 10, 'need': 0} → Avoid? ❌
Event: {'risk': 10, 'need': 95} → Avoid? ✅
Event: {'risk': 85, 'need': 0} → Avoid? ❌
Event: {'risk': 99, 'need': 95} → Avoid? ✅
Event: {'risk': 25, 'need': 85} → Avoid? ✅
This suggests that the system uses threshold-based logic for making decisions.
Why is it Relevant?
Subjectivity-IA offers an alternative to machine learning models by providing a deterministic and transparent approach to decision-making. This is especially useful in applications where explainability and traceability are crucial, such as in security systems, medicine, or computational ethics
How to Get Started?
It is currently a laboratory version that is published in https://github.com/JennyLab/Subjectivity-IA.git
Add comment
Comments