Bayesian update visualiser
Add observations and watch a prior assumption become a posterior distribution with a credible interval.
Interactive explainers
Small, fast simulations in your browser, with no sign-up or tracking scripts. Every calculation runs locally on your device.
20 simulations
No matches. Try a broader term or select all subject areas.
Add observations and watch a prior assumption become a posterior distribution with a credible interval.
Calculate the probability that B is better, expected loss and an uplift interval from visitors and conversions.
Bootstrap explained: compare mean and median, resample observations with replacement and explore a confidence interval in an interactive local example.
Understand probability calibration through a reliability diagram, Brier score and temperature correction using interactive, fully synthetic predictions.
Understand causal inference through confounding, Simpson’s paradox and standardisation in an interactive comparison of training and success rates.
Understand EVPI and EVSI through a market-test decision tree. Compare uncertainty, test accuracy and cost in an interactive example with explicit assumptions.
Bayesian optimisation explained: explore a real Gaussian process, model uncertainty and expected improvement while searching for a good parameter setting.
How should limited labour and material be shared between two products? A small production model shows how constraints shape feasible plans and why whole units can change the answer.
How much should you order when demand is uncertain? The same numbers produce different quantities under different decision rules. This calculator makes explicit which uncertainty and which kind of loss each rule addresses.
Estimate π with random points and simulate a project with uncertain task durations thousands of times.
A team can have enough average capacity and still create long queues. This simulation explores how utilisation, variable service times and a second server interact.
Three strategies, the same arms: which finds the best option sooner, and what does learning along the way cost?
A small world with a goal, traps and walls: how value iteration plans and Q-learning learns by trying.
Compare Thompson sampling with and without context. Change the context mix, number of decisions and an optional environment shift.
Calculate the optimal route with an adjustable damage penalty and step limit. Compare model score, damage and a separate audit score.
A curve can fit familiar observations extremely well and predict new ones poorly. Compare polynomial degrees on the same training and holdout data, and see when additional flexibility becomes harmful.
Which customers are similar? The answer also depends on how similarity is measured. Group the same synthetic customers by order count and monthly revenue, and explore the effect of scaling.
Select a query token, adjust temperature and masking, and inspect actual softmax weights and the calculated output vector.
Search five fictional documents with TF-IDF. Inspect scores, dates, archived rules and conflicting approval limits.
Three tasks, one machine: choose a rule preset, deadline and objective. Every start time is checked, making optimality or infeasibility inspectable.
smartlytics is the learning hub of NyxAI GmbH.
Explore the methods we work with, explained openly, with sources and calculations you can reproduce.
NyxAI is developing a Decision Core for Decision Intelligence that combines optimisation, simulation, statistics and reinforcement learning: compare options, make uncertainty visible and learn from outcomes, with people in control.
Explainable chess AI with human playing styles: NyxAI’s testing ground for understandable decisions under time pressure.