0

Is there a specific list of algorithms that tackle the bias problem well? This search doesn't seem to yield much on Google.

rahs
  • 633
  • 1
    several concepts (definitions) need to be clarified. I think here the ML means maximum likelihood. When we talk about ML, we mean MLE or LRT. They are not algorithm. For MLE and LRT, sometimes the solutions are so easy to get, and sometimes there is no analytical solution, and under this situation we need to employ the special algorithms (which is a branch of mathematics and computer science) to get the solution. Generally, the statisticians are not expert on computational mathematics. The bias is another special concept which means the deference between expected value of estimate – user158565 Dec 06 '18 at 02:22
  • 1
    and the true value of the parameter. In MLE, some of them are unbiased, some of them biased. For the biased MLE, algorithm can lot lower the biased. In addition, the bias for one time estimate has no definition and we have no base to discuss this problem. – user158565 Dec 06 '18 at 02:26

0 Answers0