Skip to main content
. 2019 Feb 28;13:135. doi: 10.3389/fnins.2019.00135

Table 1.

Overview of model pros & cons, updated from Hastie (2003).

Characteristics Neural network SVM Decision tree RF Generelized linear model Gaussian mixture model k-NN Boosting
Model complexity High High Low Fair Low High Low Fair
Sensitivity to dataraji sparsity High High Low Fair Low High High Fair
Sensitivity to data bias High High High High High High High High
Interpretability Poor Poor Fair Poor Good Poor Good Poor
Predictive power Good Good Poor Good Poor Good Poor Good
Ability to extract linear combinations of features Good Good Poor Poor Poor Poor Poor Poor
Natural handling ofraji missing values Poor Poor Good Good Poor Good Good Good
Robustness to outliers in input space Poor Poor Good Good Fair Good Good Good
Computational scalability Poor Poor Good Good Good Poor Poor Good

SVM, Support Vector Machine; RF, Random Forest; k-NN, k-Nearest Neighbors.