Linear Regression Formula Regression coefficients The regression coefficient of 𝒚 on 𝒙 is denoted by 𝒃𝒚𝒙 . 𝑏𝑦𝑥 = 𝑏𝑦𝑥 = 𝑏𝑦𝑥 = 1 𝑛 1 ∑ 𝑥 2 − (∑ 𝑥)2 𝑛 ∑ 𝑥𝑦− ∑ 𝑥 ∑ 𝑦 ∑(𝑥−𝑥̅ )(𝑦−𝑦̅) ∑(𝑥−𝑥̅ )2 1 𝑛 1 ∑ 𝑢2 − (∑ 𝑢)2 𝑛 ∑ 𝑢𝑣− ∑ 𝑢 ∑ 𝑣 (We use when 𝑥 , 𝑦 are small numbers) (When 𝑥 − 𝑥̅ , 𝑦 − 𝑦̅ are small fraction less numbers) (when 𝑢 = 𝑥 − 𝐴 , 𝑣 = 𝑦 − 𝐵 , A and B are assumed means) 𝑏𝑦𝑥 = 𝑟. 𝜎𝑦 𝜎𝑥 (Where 𝜎𝑥 is the standard deviation of 𝑥-variate, 𝜎𝑦 is the standard deviation of 𝑦-variate and 𝑟 is the coefficient of correlation) The regression coefficient of 𝒙 on 𝒚 is denoted by 𝒃𝒙𝒚 . 𝑏𝑥𝑦 = 𝑏𝑥𝑦 = 𝑏𝑥𝑦 = 1 𝑛 1 ∑ 𝑦 2 − (∑ 𝑦)2 𝑛 ∑ 𝑥𝑦− ∑ 𝑥 ∑ 𝑦 ∑(𝑥−𝑥̅ )(𝑦−𝑦̅) ∑(𝑦−𝑦̅)2 1 𝑛 1 ∑ 𝑣 2 − (∑ 𝑣)2 𝑛 ∑ 𝑢𝑣− ∑ 𝑢 ∑ 𝑣 (We use when 𝑥 , 𝑦 are small numbers) (When 𝑥 − 𝑥̅ , 𝑦 − 𝑦̅ are small fraction less numbers) (when 𝑢 = 𝑥 − 𝐴 , 𝑣 = 𝑦 − 𝐵 , A and B are assumed means) 𝑏𝑥𝑦 = 𝑟. 𝜎𝑥 𝜎𝑦 (Where 𝜎𝑥 is the standard deviation of 𝑥-variate, 𝜎𝑦 is the standard deviation of 𝑦-variate and 𝑟 is the coefficient of correlation) Tapati's Classes Co-efficient of correlation 𝒓(𝒙, 𝒚) 𝒐𝒓 𝝆(𝒙, 𝒚) 𝑟 2 = 𝑏𝑦𝑥 . 𝑏𝑥𝑦 and 0 ≤ 𝑟2 ≤ 1 𝑟 = √𝑏𝑦𝑥 . 𝑏𝑥𝑦 and −1 ≤ 𝑟 ≤ 1 𝑏𝑥𝑦 , 𝑏𝑦𝑥 𝑎𝑛𝑑 𝜌(𝑥, 𝑦) are of same sign. Equations of two lines of regression The regression equation of 𝑦 on 𝑥 is 𝑦 − 𝑦̅ = 𝑏𝑦𝑥 (𝑥 − 𝑥̅ ) The regression equation of 𝑥 on 𝑦 is 𝑥 − 𝑥̅ = 𝑏𝑥𝑦 (𝑦 − 𝑦̅) The two regression line intersect at (𝑥̅ , 𝑦̅) The acute angle 𝜽 between two regression lines is given by tan 𝜃 = | 1− 𝑟 2 𝑏𝑥𝑦 + 𝑏𝑦𝑥 | If two lines coincide then 𝜃 = 0 . So 1 − 𝑟 2 = 0 and 𝑟 = ±1 Tapati's Classes