Project checklist Flashcards
(32 cards)
1
Q
1.
A
- Frame the problem and look at the big picture
2
Q
2.
A
- Get the data
3
Q
3.
A
- Explore the data to gain insights
4
Q
4.
A
- Prepare the data
5
Q
5.
A
- Explore many different models and shortlist the best ones
6
Q
6.
A
- Fine-tune your models and combine them into a great solution
7
Q
7.
A
- Present your solution
8
Q
8.
A
- Launch, monitor, and maintain your system
9
Q
1.1
A
1.1 Understand business problem and objective, explore hidden assumptions
10
Q
1.2
A
1.2 Current solution
11
Q
1.3
A
1.3 Measure of success
12
Q
2.1
A
2.1 Get access to data and create workspace
13
Q
2.2
A
2.2 Consider legal and privacy obligations
14
Q
2.3
A
2.3 Set test data aside
15
Q
3.1
A
3.1 Check feature characteristics
16
Q
3.2
A
3.2 Visualize the data
17
Q
3.3
A
3.3 Study feature correlations
18
Q
4.1
A
4.1 Data cleaning
19
Q
4.2
A
4.2 Feature selection and engineering
20
Q
4.3
A
4.3 Feature scaling
21
Q
5.1
A
5.1 Train and compare many quick and dirty models
22
Q
5.2
A
5.2 Quick feature selection and engineering, iterate
23
Q
5.3
A
5.3 Shortlist 3-5 most promising models, preferring different errors
24
Q
6.1
A
6.1 Fine tune hyper-parameters, optimize against business objective
25
6.2
6.2 Try ensemble method
26
6.3
6.3 Measure performance against test set
27
7.1
7.1 Document
28
7.2
7.2 Explain achievement of business objective
29
7.3
7.3 Present other interesting points
30
8.1
8.1 Make production ready
31
8.2
8.2 Write monitoring code
32
8.3
8.3 Set up retraining of model
