Fraud detection on SageMaker

Step 1 - Data Preparation

I will use dataset from Kaggle that is used for Credit Card Fraud detection. In this example, there are two notebooks:

linear learner: it will use AWS SageMaker linear learner algorithm to get a base line model. The key is to use linear learning weight function and objective function to help us quickly to achieve the gaol. ref
xgboost: it will use SageMaker build in xgboost algorithm and HPO to get the model. xgboost notebook is credited to wmlba.
QuickSight for Fraud detection lab

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
0-data-preparation		0-data-preparation
1-xgboost-baseline		1-xgboost-baseline
2-linear-learner		2-linear-learner
3-auto-pilot		3-auto-pilot
4-neural-network		4-neural-network
5-xgboost-balanced-data		5-xgboost-balanced-data
6-draw-xgboost-graph		6-draw-xgboost-graph
7-auto-encoder		7-auto-encoder
99-working-notes		99-working-notes
.gitignore		.gitignore
Confusion-Matrix.pdf		Confusion-Matrix.pdf
README.md		README.md
creditcardfraud.zip		creditcardfraud.zip