Thank you very much for your great lecture. Is there any simple way, to specify how large a cluster should be? Let's say I have 20 datapoints, which I want to cluster in groups of each 5 points. Is that possible with sklearn?
I am not sure whether there are clustering algos that can produce same-sized clusters. The only one I can think of is a variation of K-means called "same-sized K-Means". It's described here: elki-project.github.io/tutorial/same-size_k_means. It would require some tweaking of the scikit-learn KMeans class to get that to work in scikit-learn though.
Thank you very much for your great lecture. Is there any simple way, to specify how large a cluster should be? Let's say I have 20 datapoints, which I want to cluster in groups of each 5 points. Is that possible with sklearn?
I am not sure whether there are clustering algos that can produce same-sized clusters. The only one I can think of is a variation of K-means called "same-sized K-Means". It's described here: elki-project.github.io/tutorial/same-size_k_means. It would require some tweaking of the scikit-learn KMeans class to get that to work in scikit-learn though.
❤❤❤❤