alexbirkett

q1 = Q( s[t], a[t] );
qMax = 0;
for (a = 0; a < A; ++a ) {
    for (s = 0; s < S; ++s ) {
        q_a = Q( s, a ) * P ( s[t], s, a );
        qMax = max( qMax, q_a);
    }

}
q2 = q1 + alpha * ( Reward(t) + D * qMax - q1 );