Curse of Dimensionality

• In an n dim hypercube → greatest possible distance is $\sqrt{n}$
• Aka the higher the dimension → wider the training points from each other
• But there are fewer data points than dimensions and the distances are huge
• Dimensionality Reduction