The curse of dimensionality

A few weeks into being a machine learning engineer, Athon told me about the curse of dimensionality. It's this thing in data science when you have a large amount of categories without many examples of each category. We came across it on a classification problem. There were as many classes as samples. We had to reduce the number of classes so it was easier to work with the samples we had. Once the classes were reduced, a model could learn the relationships between them easier.

The curse of dimensionality isn't limited to data science problems. It occurs in life too. When you have so many things on, you can't dedicate the same amount of energy to each one.

Try to focus on one thing and you'll be distracted by the other thing you have to do. Do the other thing and then the next and your list is still longer than your arm.

The cure?

The same as the in our classification problem. Reduce the number of classes. Except this time, classes are the things you've said yes to. Reduce it down to the things which matter and then you'll be able to dedicate the energy they require.