I'm looking for a way to constraint one coefficient to an interval (0,1) in a regression run by Markov switching model. Ideally, The equation enables a smoothing behaviour through the constrained coeff. {a} as following:
y = {a}*L1.y + (1 - {a})*({b}*x1 + {c}*x2 + {d})
y = {a}*L1.y + (1 - {a})*({b}*x1 + {c}*x2 + {d})
Comment