r:regression
Differences
This shows you the differences between two versions of the page.
| Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
| r:regression [2025/10/13 07:33] – [3] hkimscil | r:regression [2025/10/13 07:57] (current) – [10] hkimscil | ||
|---|---|---|---|
| Line 294: | Line 294: | ||
| </ | </ | ||
| ===== 4 ===== | ===== 4 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > ss.x <- sum((x-mean(x))^2) | > ss.x <- sum((x-mean(x))^2) | ||
| Line 307: | Line 309: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| ===== 5 ===== | ===== 5 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > y.pred <- a + b*x | > y.pred <- a + b*x | ||
| Line 356: | Line 364: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | {{: | ||
| + | </ | ||
| + | </ | ||
| ===== 6 ===== | ===== 6 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > y.hat <- mod.r$fitted.values | > y.hat <- mod.r$fitted.values | ||
| Line 520: | Line 535: | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | 아래를 실제로 구해서 | ||
| + | * y.obs 실제 y 값 | ||
| + | * y.hat prediction 값 | ||
| + | * y.mean y 평균 | ||
| + | 그 다음 residual값, | ||
| + | * res = y.obs - y.hat | ||
| + | * reg = y.hat - y.mean | ||
| + | * tot = y.obs - y.mean | ||
| + | 각각의 Sum of Square값을 구한다 | ||
| + | * ss.res <- sum(res^2) | ||
| + | * ss.reg <- sum(reg^2) | ||
| + | * ss.tot <- sum(tot^2) | ||
| + | 그 후에 각각의 df는 | ||
| + | * df.res <- n - (# of parameters (a and b) = 2) = 36 - 2 = 34 | ||
| + | * df.reg <- # of parameters - 1 = 2 - 1 = 1 | ||
| + | * df.tot <- # of observation (36) - 1 = 35 | ||
| + | </ | ||
| + | </ | ||
| ===== 7 ===== | ===== 7 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > ms.tot <- ss.tot / df.tot | > ms.tot <- ss.tot / df.tot | ||
| Line 534: | Line 572: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | <fc # | ||
| + | MS total = y의 분산값 | ||
| + | MS res = residual의 분산값 = 독립변인이 있음에도 불구하고 랜덤하게 나타난 분산 | ||
| + | MS reg = regression의 분산값 = 독립변인의 영향력으로 생긴 분산 | ||
| + | F 값은 독립변인때문에 생긴 차이값 / 랜덤 차이값 = MS reg / MS res | ||
| + | |||
| + | |||
| + | </ | ||
| + | </ | ||
| ===== 8 ===== | ===== 8 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > f.cal <- ms.reg/ | > f.cal <- ms.reg/ | ||
| Line 573: | Line 625: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| ===== 9 ===== | ===== 9 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > se.res <- sqrt(ss.res/ | > se.res <- sqrt(ss.res/ | ||
| Line 583: | Line 642: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | residuals의 (잔차들의) standard deviation (표준편차) | ||
| + | = <fc # | ||
| + | </ | ||
| + | </ | ||
| ===== 10 ===== | ===== 10 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > ss.x <- sum((x-mean(x))^2) # ss for x | > ss.x <- sum((x-mean(x))^2) # ss for x | ||
| Line 622: | Line 690: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | * b에 대한 standard error (standard deviation of b estimation) | ||
| + | * '' | ||
| + | |||
| + | * b값에 대한 significant test = | ||
| + | * b값이 (기울기가) y를 설명하는데 기여했는가? | ||
| + | * 기여했다는 가설을 테스트하는 것 | ||
| + | * 기여를 하지 않았다면 | ||
| + | * 기울기가 평균값과 같은 역할밖에 하지 못했다는 뜻이므로 | ||
| + | * '' | ||
| + | * 이 값이 t 값 (t.cal) | ||
| + | |||
| + | </ | ||
| + | </ | ||
| + | |||
| ===== 11 ===== | ===== 11 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > ################################## | > ################################## | ||
| Line 649: | Line 736: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| + | |||
| ===== 12 ===== | ===== 12 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > # Train the model with scaled features | > # Train the model with scaled features | ||
| Line 656: | Line 751: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| + | |||
| ===== 13 ===== | ===== 13 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > a <- rnorm(1) | > a <- rnorm(1) | ||
| Line 667: | Line 770: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| + | |||
| ===== 14 ===== | ===== 14 ===== | ||
| + | <WRAP group> | ||
| + | <WRAP column half> | ||
| < | < | ||
| > nlen <- 75 | > nlen <- 75 | ||
| Line 709: | Line 820: | ||
| > | > | ||
| </ | </ | ||
| + | </ | ||
| + | <WRAP column half> | ||
| + | ++++++++++++++++++++ | ||
| + | </ | ||
| + | </ | ||
| + | |||
r/regression.1760308409.txt.gz · Last modified: by hkimscil