Linear regression

GPTKB entity

Statements (50)
Predicate Object
gptkbp:instanceOf statistical analysis
gptkbp:assumes homoscedasticity
linear relationship
independence of errors
no multicollinearity
normality of errors
gptkbp:can_be_extended_to gptkb:polynomial_regression
logistic regression
gptkbp:can_be_regularized_by gptkb:lasso_regression
ridge regression
gptkbp:evaluated_by gptkb:adjusted_R-squared
mean squared error
R-squared
gptkbp:hasModel relationship between dependent and independent variables
gptkbp:hasType gptkb:multiple_linear_regression
gptkb:simple_linear_regression
https://www.w3.org/2000/01/rdf-schema#label Linear regression
gptkbp:implementedIn gptkb:SAS
gptkb:SPSS
gptkb:scikit-learn
gptkb:statsmodels
R
gptkbp:introduced gptkb:Francis_Galton
gptkbp:limitation assumes linearity
sensitive to outliers
cannot model non-linear relationships
requires large sample size for stability
gptkbp:output regression coefficients
intercept
predicted values
gptkbp:popularizedBy gptkb:Karl_Pearson
gptkbp:reduces sum of squared errors
gptkbp:relatedTo statistical analysis
ANOVA
correlation
gptkbp:requires dependent variable
independent variables
gptkbp:solvedBy gptkb:ordinary_least_squares
gradient descent
gptkbp:used_in gptkb:machine_learning
data analysis
statistics
gptkbp:usedFor feature selection
trend analysis
forecasting
predicting continuous outcomes
quantifying relationships
gptkbp:visualizes scatter plot
gptkbp:bfsParent gptkb:Method_of_Least_Squares
gptkbp:bfsLayer 6