ABSTRACT
There are many domains that machine learning can be applied to. Whilst regression and classification are the most common, there is also cost-sensitive classification, survival analysis, density estimation, cluster analysis, spatial analysis, and more. Cost-sensitive classification is a subfield of classification in which there are unequal costs associated with misclassifications. Survival analysis is concerned with making time-to-event predictions in the presence of ‘censoring’. Density estimation is an unsupervised learning task to estimate the probability density function of a variable. Cluster analysis is another unsupervised task, in which homogenous clusters of data points are estimated. Finally, spatial analysis is the supervised prediction of data with spatial features.
This chapter introduces these tasks which are implemented across mlr3proba, mlr3cluster, mlr3spatiotempcv, and mlr3spatial. A brief introduction to each is provided, as well as the key differences from classification and regression. Practical experiments are included to highlight key functionality.
