109–Data Mining Project 1_Solution

Description

9. Feature:

「Info」: 1 row represents a patient. 「共病症_Comorbidities」: A number indicates that the patient has a comorbid condition, and the label “0” means that there is no comorbid condition.

「TPR」: Time series data of “temperature, pulse, breathing rate, systolic blood pressure, diastolic blood pressure”. The same “No” means the data of the same patient.

If you want to merge the two sheets “Info” and “TPR”, you can use “No” as the merge index. Please note that “No” is the patient ID and you should not train it as a Feature otherwise your Model will learn the wrong information.

35. Submitted file name and format:

35. 1. <your-student ID>.csv

35. 1. • This file is your predict result for test data.

35. 1. • Please rename your “Submission.csv” to “<your-student ID>.csv” EX: 0760406.csv

       • 「Target」column only can be 0 or 1.

35. 1. • The entire CSV file can only have two columns.

2. <your-student ID>_Report.pdf

2. • File name: <your-student ID>_Report.pdf

   • EX: 0760406_Report.pdf

2. • Size: 12

2. • Font：Times New Roman

2. • Include:

The content must include “Data preprocessing”, “Formula”, “Validation method”.

Can be increased if necessary.

61. Submission:

    • Please upload the zip to e-Campus system “Project_01”.

IV. Rule

Items that will lose points	Project Score
Any wrong file name	-20%
Any format error	-20%
Late project per day	-20%

	Item	Total score for the entire semester
	F1_score		10%
	(base on Target=1)
	Project Report		10%
	*−=∗		( ∗ )		=
			( + )			( + + )

TP: True Postive

TN: True Negetive

FP: False Postive

22. Deadline

– 2020/11/13 (Fri.)

PM 11:50 upload deadline.