For each following case,
* Identify important features (5)
* Draw AUC-ROC curve (5)
* Interpret results (5)
Cases:
1. Use logistic regression to regress news categories on (latent) components
2. Use logistic regression to regress senders' affiliation on (latent) components
* academia vs. industry (you can identify it by email address domain name)
3. Use logistic regression to regress receivers' affiliation on (latent) components
4. Use decision tree to do the same job as cases 1,2,3 (draw decision trees) (extra 10)
5. Change the #top words to include and #components; repeat 1,2,3,4. (extra 10)
'설명없음' 카테고리의 다른 글
| [StrataScratch] Highest Cost Orders [MySQL] (0) | 2022.06.26 |
|---|---|
| 백준 2961: 도영이가 만든 맛있는 음식 (0) | 2022.06.16 |
| 백준 2295: 세 수의 합 [C/C++] (0) | 2022.06.16 |
| 백준 7453: 합이 0인 네 정수 [C/C++] (0) | 2022.06.13 |
| 백준 2636: 치즈 [C/C++], 삼성 코딩 테스트 (0) | 2022.06.13 |