Dummy

BayesHive
BayesHive is a web application for data analysis using powerful Bayesian statistical methods. We help you build statistical models for your entire dataset such that final decisions and measurements are based on all the available information.


More Videos

Getting started (12m17s)
Dynamical systems (36m05s)
Coming soon: Bayesian statistics

Solutions for

Work directly with stochastic financial models
  • any model based on stochastic differential equations
  • including models with unobserved variables and timeseries
  • estimate parameters from historical data
  • Outcomes: probabilistic portfolio forecasts, derivatives pricing, risk estimates
Use BayesHive to analyse your data
  • Linear and nonlinear regression, repeated measures ANOVA
  • Bayesian effect size estimates
  • Dynamical systems models for physics, biosignals and behaviour
  • Collect analyses into a printable document
  • Share your models, data and documents
Using BayesHive is the easiest way to learn about:
  • Statistical modeling
  • Probability distributions
  • Uncertainty and risk
  • Bayesian statistics
  • Dynamical systems


How it works

Upload data

  • Spreadsheets
  • Time series
  • MATLAB .MAT files
  • Quandl data
  • or use shared data

Build statistical models

Using
  • Probability distributions
  • Dynamical systems
  • Stochastic differential equations
Through a point-and-click web interface or
  • run shared models

In order to

  • Estimate effect sizes
  • Measure parameters
  • Quantify uncertainty
  • Predict and forecast
  • Test hypotheses
  • Calculate risk
  • Make optimal decisions

What is...?

Bayesian statistics provides the mechanics for learning from real-world data based on statistical models that can be simple or very complex. Because Bayesian statistics is based exclusively on the laws of probability, it is easy to understand but very powerful; when applied correctly, it leads to optimal decisions.
Wikipedia | Scholarpedia | Sivia | Gelman et al.
In Fully Bayesian computing one programs directly with probability distributions to build a statistical model for observed data. Everything else becomes trivial — and provably optimal. That includes decisions, forecasts, uncertainty quantification, risk assessments, control algorithms and measurements of underlying states.
Kerman and Gelman | Wired | Beau Cronin | Microsoft
Baysig is a new probabilistic programming language that combines modern programming languages with direct support for Bayesian inference and dynamical systems. BayesHive writes Baysig code for you to support data analysis.
Quick tour | More help