def upper_bound(self,t):
mu = np.true_divide(self.success,self.trials)
interval = np.sqrt(self.alpha*np.log(t)/(2.0*self.trials))
return mu+interval
微信
公众号
扫码关注公众号