Author : Amritha Varma
Locally Estimated Scatterplot Smoothing (LOESS) is a regression tool which help us to create a smooth line between the scatter plot. It helps us to show the relation between the variables and trends of variables. It is a non parametric regression method which combines multiple regression in K-nearest neighbor. Non parametric regression finds a curve without assuming data. This smoothing function captures general patterns and it makes assumptions about relationship among variables. The result of LOESS is a line moving through central tendency. It is mainly used to show the relationship between two variables with large data sets.
Content:-
What is regression?
What is smoothing?
Uses.
Advantages.
Disadvantages.
Example.
Reference.
WHAT IS REGRESSION:-
Regression is a mathematical function which show the relationship between one dependent variable and one more variable. The obtained function is called regression equation.
WHAT IS SMOOTHING:-
Smoothing is a technique to group variables with similar expectations and fit a suitable curve. It helps to decrease the volatility in data series. Therefore trend can be observed clearly.
USES:-
Fitting a line to a scatter plot where noisy data values with your ability to see a line of best fit.
Linear regression where least squares fitting does not create a line of good fit
Data explorations in social science.
ADVANTAGES:-
Easy to use
Simple
Flexible
Shows trend
DISADVANTAGES:-
Complex in nature
Difficult for explaining result
No on hand formula so it is difficult to transport results.
EXAMPLE:-
Figure-1
Here the scatterplot will be displayed on the basis of car dataset with box plots in the margins and non-parametric regression smooth.
x <- mtcars$wt
y <- mtcars$mpg
plot(x, y, main = "Main title",
xlab = "X axis title", ylab = "Y axis title",
pch = 19, frame = FALSE)
lines(lowess(x, y), col = "green")
Commentaires