The Impact of Behavioral Biases to Real Data Science Projects
summary
Describes common human and managerial behavioural biases that derail real-world data science projects—such as overconfidence in outcomes, underestimating uncertainty, and misaligned expectations from leaders transitioning from IT/software delivery—and explains how these biases affect choices about data collection, model interpretability, maintainability, integration, and ROI. Uses the speaker’s ~20 years of experience in startups, R&D and consulting to give concrete examples of these pitfalls and practical suggestions for identifying and mitigating them in organizational processes and project planning.
AI-written summary · about
chapters (6)
- 0:00Start
- 9:08Behavioral bias : The certainty preference
- 10:56Pitfall : Failure to Baseline
- 14:24Behavioral bias : The bricklayer mindset
- 15:46Pitfall: The Problem Formulation Hurdle
- 27:18Pitfall: The causation vs correlation trap
related talks
01h 00mCognitive biases in software development
30mBehavioral Software Engineering
39mHow Cognitive Biases Affect our Software Architectures
44mCognitive Biases Driven Architecture?
01h 02mCognitive biases in software development
01h 00mUnbiasing teams
11mHow to deal with your demons
52mYour Brain on Software Development