Active Learning of Nondeterministic Finite State Machines

Table 3

Processing the counterexample 𝑎/𝑦⋅𝑏/𝑦⋅𝑎/𝑦⋅𝑏/𝑥  for 𝑀₀⁽¹⁾.

(a) Adding the suffixes of 𝑎⋅𝑏 to 𝐸

𝑇2	𝐸
	𝑎	𝑏	𝑎⋅𝑏
𝑆
𝜖	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦	𝑦	𝑥,𝑦	𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑥	{𝑥}	{𝑥}	{𝑥⋅𝑥}
𝑆⋅𝐼/𝑂
𝑏/𝑦	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦⋅𝑎/𝑦	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑦	𝑦	𝑦	𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑥⋅𝑎/𝑥	{𝑥}	{𝑥}	{𝑥⋅𝑥}
𝑎/𝑦⋅𝑏/𝑥⋅𝑏/𝑥	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦

(b) Moving the rows 𝑎/𝑦⋅𝑏/y to 𝑆

𝑇2	𝐸
	𝑎	𝑏	𝑎⋅𝑏
𝑆
𝜖	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦	𝑦	𝑥,𝑦	𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑥	{𝑥}	{𝑥}	{𝑥⋅𝑥}
𝑎/𝑦⋅𝑏/𝑦	𝑦	𝑦	𝑦⋅𝑦
𝑆⋅𝐼/𝑂
𝑏/𝑦	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦⋅𝑎/𝑦	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑥⋅𝑎/𝑥	{𝑥}	{𝑥}	{𝑥⋅𝑥}
𝑎/𝑦⋅𝑏/𝑥⋅𝑏/𝑥	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑦⋅𝑎/𝑦	𝑦	𝑦	𝑦⋅𝑦
𝑎/𝑦⋅𝑏/𝑦⋅𝑏/𝑦	𝑦	𝑦	𝑦⋅𝑥,𝑦⋅𝑦