First of all: I am a non-statistician who's currently trying to dive deeper into the topic of Data Science, so this question might sound really naive. So, I was wondering about the major distinction aspects between Machine Learning and Statistics / Statistical Modeling. Although there exist many aspects, in which both disciplines appear to differ, many sources mentioned that its corresponding purpose as main aspect: While Statistics majorily aims to make inferences about relationships between variables, machine learning rather aims to achieve a high accuracy for future predictions. While this is still easy to comprehend, I am getting confused if this is not just two sides of the same coin. Isn't a high accuracy of a model regarding future predictions an indicator for its generalization ability? What is statistics doing that machine learning doesn't, that makes its output still more inferentiable although accuracy is noticeably lower?

I highly appreciate your support!



Browse other questions tagged or ask your own question.