Big-data machine learning has emerged as an important research topic because data larger than a machines capacity are now very common.However,there are many challenges in applying bigdata machine learning.First,most traditional machine learning algorithms were designed to run on a single computer.Second,sub-sampling data to one machine for analysis is always an option,so when distributed machine learning is preferable is an issue.After an overview of these challenges,we discuss distributed machine learning methods.In particular,a Newton method for large-scale logistic regression is given as an example.We also briefy review other existing developments and show how they are used in practical applications.Finally,we argue that big-data machine learning involves many issues other than algorithms.Programming environments,systems,and the application workfow must be considered together for a successful big-data machine learning project.