Sampling Bias Correction for Supervised Machine Learning: A Bayesian Inference Approach with Practical Applications
This provides practitioners in fields like medical sciences, image recognition, and marketing with tools to correct sampling bias in their inference pipelines, though it is incremental as it adapts existing Bayesian methods to a specific problem.
The paper tackles the problem of training supervised machine learning models when the training set has a known sampling bias, achieving this by using a Bayesian inference framework to adjust the posterior distribution for the sampling function, with applications in binary logistic regression and scenarios like label imbalance.
Given a supervised machine learning problem where the training set has been subject to a known sampling bias, how can a model be trained to fit the original dataset? We achieve this through the Bayesian inference framework by altering the posterior distribution to account for the sampling function. We then apply this solution to binary logistic regression, and discuss scenarios where a dataset might be subject to intentional sample bias such as label imbalance. This technique is widely applicable for statistical inference on big data, from the medical sciences to image recognition to marketing. Familiarity with it will give the practitioner tools to improve their inference pipeline from data collection to model selection.