Maximum Likelihood Inference
Göran Kauermann, Helmut Küchenhoff, Christian Heumann · Springer series in statistics · 2021
In the last chapter, we explored Maximum Likelihood as a common approach for estimating model parameters given a sample from the population. In this chapter, we examine in more detail this estimate as well as the likelihood and its related functions: the log-likelihood, score function and Fisher information. We also derive properties of their asymptotic distributions, which tells us how much the functions themselves vary between samples. This is useful for deriving confidence intervals and statistical tests. We begin this chapter with a peek into the controversial history of the likelihood function. Not only does this give an insight into the often fraught process of discovery, but also makes clear the sometimes confusing difference between likelihood and probability.
Cited by 4 works. doi:10.1007/978-3-030-69827-0_4
Authors
References
- A Stochastic Approximation Method
- On the mathematical foundations of theoretical statistics
- 001: On an Absolute Criterion for Fitting Frequency Curves.
- Practical Markov Chain Monte Carlo
- The Messenger of Mathematics
- R.A. Fisher and the making of maximum likelihood 1912-1922
- The History of Likelihood
- The Epic Story of Maximum Likelihood