Statistical Simulation

Sometimes it's not possible to use classical configurations to make inferences about things, e.g. our sample size is small

when the traditional methods don't work, but the know the truth about the word and can control it though computerized stimulations

When (Nearly) Normal models don't work

Sometimes Statistical Inference can be done without applying theoretical models, but instead with using brute force: generating the data ourselves.



Links:


Sources

Machine Learning Bookcamp: Learn machine learning by doing projects. Get 40% off with code "grigorevpc".

Share your opinion