 How can we mitigate the curse of dimensionality? So the curse of dimensionality is a problem in many fields of mathematics where as the number of dimensions increase, the sense of relative distances vanishes. So in a very high dimensional space, the distance A, B, and A, C become indistinguishable. And this could lead to problems such as overfitting in the field of machine learning. One way to deal with this problem is to get more training data. This will pad the space and thus reintroduce the sense of relative distances and neighborhoods. However, you typically need an exponential amount of data in order to curb the curse of dimensionality. And so another way to curb this curse is to do dimensionality reduction by projecting all of these points on a lower dimensional space and then performing some post-processing or machine learning after that.