ML Challenges

• Insufficient data - Small data sets often suffer from sampling noise.
• Non-representative data - Large data sets can suffer from sampling bias.
• Poor-quality data
• Irrelevant features
• Overfitting
• Underfitting

References