Probability & Statistics
Uncertainty, Bayes, Gaussians — the basis of estimation.
Why it matters in robotics
Nearly every robotics estimation system — localization, SLAM, sensor fusion, tracking — is built on Bayesian reasoning over noisy, uncertain data, so interviewers use probability as a litmus test for whether you can reason about state estimation at all. Expect to derive Bayes' rule, manipulate Gaussians (mean/covariance, marginalization, conditioning), and explain how a prior combines with a likelihood to form a posterior. Fluency here is the prerequisite for talking credibly about Kalman and particle filters, which are common follow-up questions.
Application focus
The same topic, tailored to the robot you're building. Your choice is remembered across the roadmap and every topic.
At a glance
The Bayes filter loop: a prior belief is pushed forward by the motion model (predict), then corrected by a noisy measurement via Bayes' rule (update), producing the posterior that becomes the next prior.
What to study
- ✓Bayes' rule: priors, likelihoods, posteriors, and the normalizing evidence term
- ✓Gaussian distributions: mean/covariance, marginalization, conditioning, and why products of Gaussians stay Gaussian
- ✓Random variables, expectation, variance, and conditional/joint distributions
- ✓The Bayes filter: recursive predict (motion) and update (measurement) state estimation
Study by time budget
Pick the path that fits the time you have before your interview.