Skip to content

Myhill nerode

Myhill–Nerode Theorem: Short Notes

The Myhill–Nerode theorem provides a characterization of regular languages. It states that a language \(L\) is regular if and only if the number of equivalence classes of the relation \(\equiv_L\) (where \(x \equiv_L y\) if for all \(z\), \(xz \in L \iff yz \in L\)) is finite. This theorem is useful for proving whether a language is regular and for minimizing deterministic finite automata (DFA).

Example

Consider the language \(L = \{ w \in \{0,1\}^* \mid w \text{ ends with } 01 \}\).

DFA Transition Table

State Input 0 Input 1
\(q_0\) \(q_1\) \(q_2\)
\(q_1\) \(q_3\) \(q_4\)
\(q_2\) \(q_4\) \(q_3\)
\(q_3\) \(q_5\) \(q_5\)
\(q_4\) \(q_5\) \(q_5\)
\(q_5\) \(q_5\) \(q_5\)

Each state represents an equivalence class under \(\equiv_L\). Since there are finitely many states, \(L\) is regular.


References