✳flâneur — a map of the web's best reading

Exercises 2 -- Analytical Regression | 6.390 | IntroML | Fall24

introml.mit.edu · 28 words · saved by 1 readers

Squared error is frequently used as a loss function for regression. Which of the following pictures illustrates the squared loss function? Assume that the dark blue line is described by 𝜃 , 𝜃 0 , the black dots are the ( 𝑥 , 𝑦 ) data, and the light lines indicate the errors we are measuring (prior to squaring the error). Consider the data set and regression lines in the plot below. What is the squared error of each of the points with respect to the blue line? Basic linear regression seeks to minimize the mean squared error over all training points: In the following questions, ignore the factor of 1 / 𝑛 and consider just the terms inside the sum. What is the contribution from each point to the gradient of the objective with respect to the parameters 𝜃 and 𝜃 0 of the blue (lower) line? (Hint: re-express the above equation in terms of 𝑦 ^ = 𝜃 𝑇 𝑥 + 𝜃 0 , the predicted value.) What is the squared error of each of the points with respect to the green line? Consider a o

Exercises 2 -- Analytical Regression | 6.390 | IntroML | Fall24 Introduction to Machine Learning this page (from a previous semester) is no longer accessible, except for staff

Explore this link on the map →

related reading