Inference vs. Prediction: An Overview

Subscribe to RichardOnData here: https://www.youtube.com/channel/UCKPyg5gsnt6h0aA8EBw3i6A?sub_confirmation=1

In this video I go over the difference between inference and prediction, in the statistical modeling and machine learning context.

It happens all the time - clients have requests to incorporate machine learning and/or statistical modeling into their projects, but it is not immediately clear whether the intention is prediction, or inference. This should be one of the first questions the modeler asks. In the case of prediction, the goal is to predict future values of some outcome. In the case of inference, the goal is to best estimate how the data influence the outcome.

However, there is a trade-off here: the most predictive models are usually not the most inferential, and vice-versa. If one thinks of some kind of regression model, this is a model with great inferential capability but that tends to be outperformed by advanced machine learning methods. One can introduce regularization, in an attempt to introduce bias but decrease error -- hence, more predictive power at the expense of explainability. Then one can use algorithms like KNN, trees, or random forests, but these are very hard to explain. Neural networks are called "black box" approaches, but they have tremendous predictive power in events of large data.

#DataScience #InferenceVsPrediction #PredictionVsInference

PayPal: richardondata@gmail.com
Patreon: https://www.patreon.com/richardondata
BTC: 3LM5d1vibhp1F7pcxAFX8Ys1DM6XLUoNVL
ETH: 0x3CfC599C4c1040963B644780a0E62d45999bE9D8
LTC: MH8yPjvSmKvpmRRmufofjRB9hnRAFHfx32

Видео Inference vs. Prediction: An Overview канала RichardOnData