Did you reduce the dimension before applying HDBSCAN? | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

maCDzP on Sept 10, 2024 | parent | context | favorite | on: Show HN: I mapped HN's favorite books with GPT-4o

Did you reduce the dimension before applying HDBSCAN?

pmaze on Sept 10, 2024 [–]

I did, there was a first round of UMAP to 50 dimensions. Running HDBSCAN on the full embeddings gave bad results, lots of singleton clusters.

maCDzP on Sept 18, 2024 | [–]

Interesting, I got the opposite result. The full embeddings gave two or three clusters. How did you work with the hyper parameters of HDBSCAN?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact