Evaluating the performance of clustering algorithms

So far, we have built different clustering algorithms, but haven't measured their performance. In supervised learning, the predicted values with the original labels are compared to calculate their accuracy. In contrast, in unsupervised learning, we have no labels, so we need to find a way to measure the performance of our algorithms.