3. Consider a DFA MDm= having a set of multiple (one or more) final states (FQ) and a DFA MDs=Q,,,q0,qf> with a single final state (qfQ). [5+10=15] We now pose the following question: Is it always possible to convert an MDm to an MDs such that L(MDs)=L(MDm) ? (a) If it is possible, prove it, If it is not possible, provide a counter example. (b) If the above conversion is not possible for all regular languages, provide the characterization of the subset of regular languages where it is possible and prove it