Great work!

Data Privacy and Anonymization in Python

Rebeca Gonzalez

Data engineer

You have completed the course

Recap: What you have learned

• Sensitive and non-sensitive personally identifiable information (PII)
• Quasi-identifiers
• Linkage attacks
• Data suppression
• Data masking
• Data generalization
• Synthetic data generating
• Sampling from probability distributions for different type of attributes

Privacy models: k-anonymity

• K-anonymous datasets
• Exploring possible combinations in the dataset
• Generalizing data using hierarchies and ranges
• Avoid re-identification attacks
• Without falsifying or randomizing data!

Privacy models: differential privacy

• Differential privacy systems can measure and quantify privacy in data releases
• One of the most important definitions of privacy in present time

Differentially private models and operations

• People are increasingly working with differentially private machine and deep learning models
• Trained and run different type of differentially private machine learning models!
• Practiced advanced concepts such as privacy budget and tracking

Other interesting libraries

$$ $$

• Google's differential privacy
• TensorFlow Privacy
• ARX Data Anonymization Tool

Congrats!

Data Privacy and Anonymization in Python