REGRESSION Flashcards
what is linear regression?
shows a linear relationship between a dependent (y) and one or more independent (x) variables
it finds how the value of the dependent variable is changing according to the value of the independent variables.
what is best fit line?
When working with linear regression, our main goal is to find the best fit line that means the error between predicted values and actual values should be minimized. The best fit line will have the least error.
what is the key point of linear regression?
dependent variable must be a continuous/real value.
what are types of linear regression?
simple and multiple
what is simple linear regression?
response variable is affected by one predictor variable
what is multiple linear regression?
response variable is affected by more than one predictor variable
formula for simple linear regression?
y’=b+w1x1
formula for multiple linear regression?
y1=b+w1x1+w2x2+w3x3
if the model’s prediction is perfect then?
If the model’s prediction is perfect, the loss is zero; otherwise, the loss is greater.
what is cost function?
the different values of weights give different line of regression, so cost function is used to estimate the values of the co-efficient for the best fit line.
cost function optimizes the weights and measures how well a linear regression model is performing
ridge regression is used to prevent?
overfitting
what is regularization?
it is a strategy used to overcome overfitting by providing the machine with new knowledge. It is a type of regression. But here the co-efficient values are reduced to zero.
why use regularization?
it is not uncommon for our model to function well enough on training data but poorly on unseen or test data. the model is unable to forecast the output for unknown data and so the model is referred to as an overfitted model. To overcome this we use regularization
what is noise?
points in the data that are there by accident. They do not show the true nature of the model and can mislead.
how is cost function different in ridge and lasso regression?
The cost function for both ridge and lasso regression are similar. However, ridge regression takes the square of the coefficients and lasso takes the magnitude.
