Reading some of the online links related to interview questions for data scientist position , i stumbled upon a question – what is the difference between Machine Learning and Econometrics? Coming from economics background one always feels that there is a lot of similarity between the two but the discussion below provides a good overview

Varian(2014), Diebold(2016) and Mullainathan & Spiess (2017) have written blog posts and articles which discuss this question in detail. Following posts discusses a general theme and readers curious to learn more should definitely read the articles referenced here.

1. Machine learning is mostly concerned with both summarizing data and prediction (i.e $widehat{y}$) whereas Econometrics can be broken down into four main categories prediction, summarization, estimation and inference.
2. Machine learning is concerned with non-causal prediction whereas in econometrics we are ore concerned with causal prediction. Diebold(2016) in his posts explains machine learning non-causal prediction as “if a new person $mathrm{textit{i}}$  arrives with covariates ${X_i}$ what is my minimum – Mean Squared Error guess of ${y_i}$? ”  whereas causal prediction is more concerned with, if i intervene and give a person $mathrm{textit{i}}$ a certain treatment what is my minimum- Mean Squared Error guess of  $Delta {y_i}$ ?
3. Machine learning often uses the terms features or predictors which in econometrics is known as independent variables.

## References:

1. 2014. “Big Data: New Tricks for Econometrics.” Journal of Economic Perspectives28 (2): 3-28.  Can be downloaded here.
2. Diebold : https://fxdiebold.blogspot.com/2016/10/machine-learning-vs-econometrics-i.html
3. 2017. “Machine Learning: An Applied Econometric Approach.”Journal of Economic Perspectives31 (2): 87-106