[I'm done!] assignment 1

This is my solution for assignment1 in STA9083 class.

Please do *NOT* trust my solution entirely, and just use it as referernce, since it is my own solution and *NOT VERIFIED* 

 

If you have any comment on my solution, please leave it. Any comments are welcome!

 

 

 

이번 학기는 assignment 지옥인가요..?

 

무슨 도장 깨기도 아니고, 하나 치우면, 어느새 새로운 놈 등장..! 

너무 가혹해요.. 흑흑

 

게다가 exam과 project 도 있다는 사실..!

 

살려주세요.

 

내 주특기 발동할 시점이 되었구나.

내 주특기.. "징징거리기" ㅎㅎ

슬슬 또 "힘들다고, 대학원 괜히 왔다고" 징~징~ 징글벨 울릴 때가 되었네. ㅎ

 

푸는 중임.

이거 수식 편집은 또 어쩔거..ㅎㅎ 

---

Problem 1 [Randomized experiments]

(a) (5.2) SRE vs CRE [Done!]

---

(b) (6.7) Regression adjustment using linear regression with interaction [Done!]

---

Problem 2

(a) (10.2) Nonparametric identification of the distributional and quantile causal effects

quantile causal effect?

?

---

(b) (10.3) Outcome imputation estimator in the fully interacted logistic model [Done!]

---

(c) (11.3) Lack of invariance for the IPW estimator [Done!]

---

(d) (12.2) An alternative form of the doubly robust estimator [Done!]

---

Problem 3 [Done!]

---

---

 

 

---

Problem 4 [Almost Done, but I have to review and add more interpretation!]

---

Problem 5 (Almost done, but I have to revise interpretation, since the ATT results are quite different with ATE results!)

★ Need to revise interpretation! ★

 

---