Setosa blog
markov-chainsprobabilityvisualizationstatistics
Abstraction: Interactive visual introduction to Markov chains and transition matrices
Key points:
- Markov chains model systems hopping between states with fixed transition probabilities
- Transition matrix encodes all state-to-state probabilities; rows must sum to 1
- Number of matrix cells grows quadratically as states are added
- "Stickiness" in real-world sequences (e.g. weather) modeled by high self-transition probabilities (0.9)
- Google PageRank is a real-world example of a large Markov chain
- Named after Andrey Markov; used in meteorology, ecology, finance, computer science
Connections: Markov Chains · Probability · Stochastic Processes · Google
Source: http://setosa.io/blog/2014/07/26/markov-chains/index.html