1: \begin{abstract}%

2: Random walks are a fundamental model in applied mathematics and are a common example of a Markov chain. 

3: The limiting stationary distribution of the Markov

4: chain represents the fraction of the

5: time spent in each state during the stochastic process.  

6: A standard way to compute this distribution for a random walk on a finite set of states

7: is to compute the Perron vector of the associated transition matrix.  

8: There are algebraic analogues of this Perron vector in terms of transition probability 

9: tensors of higher-order Markov chains. 

10: These vectors are nonnegative, have dimension equal to the dimension of the state space, and sum

11: to one and are derived by making an algebraic substitution in the 

12: equation for the joint-stationary distribution of a higher-order Markov chains. 

13: Here, we present the \emph{spacey random walk}, a non-Markovian

14: stochastic process whose stationary distribution is given by the tensor

15: eigenvector.  The process itself is a vertex-reinforced random walk, and its discrete

16: dynamics are related to a continuous dynamical system.  We analyze the

17: convergence properties of these dynamics and discuss numerical methods for

18: computing the stationary distribution.  Finally, we provide several applications

19: of the spacey random walk model in population genetics, ranking, and clustering data, 

20: and we use the process to analyze taxi trajectory

21: data in New York. This example shows definite non-Markovian structure. 

22: \end{abstract}

23: