linear regret

Question 1

how to get a chi squared value

Answer 1

sum of i’s [ ((D_i - M(x_i, theta)) / sigma_Di)^2 ]
(where D_i are data points, M is the model, sigma_Di is uncertainty for D_i)

Question 2

how to get k in the chi squared distribution

Answer 2

degrees of freedom = no. of data points - no. of model parameters