The easiest way to prove that two FSM's are equivalent is to minimize both and see if the results are the same.
To do this, you can use the partitioning minimization procedure as explained in section 8.6.1 of Fundamentals of Digital Logic.
Without going into too much detail, I will illustrate this procedure for your left state machine. The idea is to partition all states in such a way that states in the same partition are equivalent.
We start by assuming that all states may be equivalent by partitioning the states into one partition. Before doing this, note that states B and F are unreachable and can be deleted from the FSM. Therefore, the first partition is as follows:
$$P_1 = (ACDE)$$
The next step is to create groups of states with the same output value. This gives the following partition:
$$P_2 = (ACE)(D)$$
Next, we have to make sure that all the 0-successors of the states in a group are contained in one group (the same holds for the 1-successors). For example, the 0-successors of (ACE) are (DAA) which are not contained in a single group. Therefore, the group (ACE) needs to be split:
$$P_3 = (A)(CE)(D)$$
Since both the 0-successors (AA) and the 1-successors (EC) of (CE) are contained within one group, this partition does not need to be split anymore. This final partition learns us that states C and E are equivalent and we need three states in the minimal FSM.
Now we can introduce names for the new states: S1 = (A), S2 = (CE) and S3 = (D) and construct a new state table:
-------------------------
Current Next state Output
State 0 1
-------------------------
S1 S3 S2 0
S2 S1 S2 0
S3 S2 S1 1
-------------------------
Doing the same for your right state machine gives the following partition:
$$P = (ACE)(D)(BF)$$
Using the state assignment S1 = (ACE), S2 = (D) and S3 = (BF) gives exactly the same state table as above. Hence, the two state machines are equivalent.
$$\blacksquare$$
If we look at the top left corner, where a=0 and c=0, we see that s follows b. This gives us a partial result of s=b. Moving down to the bottom left corner, we see that s is inverted when a=1. This is the XOR operation, and gives us s=a^b. Moving over to the right side, we see that the output is inverted when c=1. This gives us our final result of s=a^b^c.
Best Answer
A state-transition table is very similar to a truth table for combinational logic. On the input side you have a column for each bit in the state encoding and for all of the inputs. So you need to pick a state encoding and understand what the other inputs will be. Your table must then have a row for every possible combination of the current state value and the input signals, although you can use "don't care" entries for the inputs in some cases.
On the output side of the state-transition table you show the required value of the state bits for the desired next state. At that point you can use the usual techniques for optimizing the logic implementation.