Dear Asfandiyar,

Normally, you take the biggest distance, here is about 1500 and take 20% of it, or 300 in this case. Now from the top of the dendrogram measure the distance between the upper horizontal line and the second one - this gives approximately 450>300. This means you can consider the two-cluster solution for this sample. Then take the distance between the second and the third horizontal line -now we have 330 > 300, so the 3-cluster solution is also valuable. Further the distance between the third and the next horizontal line is about 280

Dear Asfandiyar Khan . See the following useful links: https://towardsdatascience.com/cheat-sheet-to-implementing-7-methods-for-selecting-optimal-number-of-clusters-in-python-898241e1d6ad#:~:text=As%20shown%20in%20Figure%206,the%20agglomerative%20hierarchical%20as%20well.

https://www.displayr.com/what-is-dendrogram/#:~:text=In%20the%20example%20above%2C%20the,between%20two%20and%20three%20clusters.

https://www.statisticshowto.com/hierarchical-clustering/

https://m.youtube.com/watch?v=ijUMKMC4f9I

Kindly see also the following useful links: https://www.datanovia.com/en/lessons/determining-the-optimal-number-of-clusters-3-must-know-methods/

https://www.geeksforgeeks.org/determining-the-number-of-clusters-in-data-mining/

https://www.analyticsvidhya.com/blog/2021/05/k-mean-getting-the-optimal-number-of-clusters/

https://academic.oup.com/ije/article/47/1/321/4091562

https://hlab.stanford.edu/brian/number_of_clusters_.html

At different levels of similarity, a different number of clusters is obtained. At the highest level of similarity, there are as many clusters as there are objects before clustering. At the lowest similarity level, 2 clusters.

It is nicely explained with a good detail here : https://www.youtube.com/watch?v=3BzfOLnIY9w and

https://www.youtube.com/watch?v=ijUMKMC4f9I