non-parametric method
A statistical or machine learning method that does not assume a fixed parametric form for the underlying distribution or function. Common in fields like density estimation, regression, and classification.
Non-parametric methods are highly flexible because they do not assume a distribution for the data.
Kernel density estimation is one of the typical non-parametric methods.
Loanword from English 'non-parametric method', combining 'non-parametric' (ノンパラメトリック) and 'method' (法).