Inferring Causality with Data from Personal Software Process
Abstract
Can Causal Algorithms Help? To control software development, we need factors that 1) Can be selected or manipulated 2) Have a causal effect (direct or indirect) on desired outcomes. New algorithms and techniques are becoming available. They are related to but distinct from multiple regression, Bayesian networks, and Machine Learning. The methods have been successful in other domains. How can we gain confidence when applying to Software Engineering?
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2018
- Accession Number
- AD1086994
Entities
People
- Michael Konrad
- William R. Nichols
Organizations
- Carnegie Mellon University