← Foundations of Mathematical Biology

Deterministic vs stochastic models

Biological systems can be represented either without explicit randomness or with randomness included as part of the model. This gives the distinction between deterministic and stochastic models.

Deterministic models

In a deterministic model, the initial conditions and parameter values determine a single model trajectory. Repeating the calculation with exactly the same inputs gives the same result.

For example,

\[\frac{dN}{dt}=rN,\qquad N(0)=N_0\]

has a uniquely determined solution

\[N(t)=N_0e^{rt}.\]

Stochastic models

A stochastic model includes random quantities or random events. Starting from the same initial state can therefore produce different possible trajectories.

Instead of asking only for one trajectory, stochastic modelling may investigate quantities such as

\[P(\text{extinction before time }T),\qquad E[N(t)],\qquad \operatorname{Var}(N(t)).\]

Comparison

DeterministicStochastic
Same inputs give the same trajectorySame inputs can give different trajectories
Describes a specified systematic evolutionRepresents variability explicitly
Often studied with difference or differential equationsOften studied with Markov models, stochastic simulation or SDEs

Why both are useful

Stochastic is not automatically better than deterministic. The appropriate choice depends on the biological question, system size, available information and importance of variability. Deterministic models can describe dominant or average behaviour efficiently, while stochastic models are especially useful when the distribution of possible outcomes matters.

Key idea. Deterministic models describe one outcome for specified inputs; stochastic models describe uncertainty by allowing multiple possible outcomes and assigning probabilities to them.