This is a set of Python Codes related to a basic data science Probability Course at a masters level. We study Estimators , in particular, MLE Estimators of various standard probability distributions.
The theory for Maximum Likelihood Estimators can be found briefly here: https://en.wikipedia.org/wiki/Maximum_likelihood_estimation
We study the various MLE Estimators for Standard Distributions. We do this by:
STEP 1 : Creating a data frame by generating data according to the probability distribution.
STEP 2 : Then we apply the MLE Estimators to make an estimate of the parameter in consideration.
STEP 3 : (optional) We estimate percentage error by checking against our known parameter. We can also visualize the data with graphs.
- Instructors can use the data generated in STEP 1 to create exercises for students to test their code.
- Sample Worksheet for such exercise is included. This can be used as a template to create further such exercises.
- The corresponding Data for this exercise is also included. The Data can be generated as required.
- Instructors can use the complete code to check the working of students code by standardizing estimation.
- Lesson notes for the theory are included: Likelihood and MLE is explained. Fisher Information is also tackled in a comparatively 'intutive' manner.
A challenge distribution "The Special Dice" is also included. The Special Dice is a dice which has a probability p of showing 1 and remaining numbers are equally probable. A challenge is to create an MLE Estimator for this parameter p and estimate it using Python code.