Discussion on "Earth Engine Tutorial #31: Machine Learning with Earth Engine - Unsupervised Classification"

Qiusheng Wu · 2020-08-07T12:43:56.344Z

This tutorial shows you how to perform unsupervised classification (e.g., KMeans clustering) in Earth Engine. Unsupervised classification algorithms available in Earth Engine Source: https://developers.google.com/earth-engine/clustering The ee.Cluste...

Professor,

Thank you for this excellent post. To your knowledge, is it possible to do fuzzy clustering (soft K-means) in addition to the classic K-means clustering with the Google Earth Engine algorithms?

Hi Qiusheng, thanks for this article, it is very useful. Its great that we can use now ee in proper Python environment ;) ...I have two questions. Q1: the centroids of the each clusters (let say NDVI values) are always in an ascending order? Q2: how can we get the centroid values from the each cluster? Thank you ;)

I don't understand your Q1. For Q2, you can convert raster to vector to get centroids.

Hi Qiusheng Wu, Thank you for your prompt reply. In Q1 I meant that, let say that I have an image and calculated the NDVI indexes. I choose 3 clusters. NDVI values will be between 0.1 and 0.7 in my image. The higher values are representing high vegetation, the medium values low vegetation, and the lower values other land types. Let's assume that the cluster centroids will be around 0.2, 0.4 and 0.6. My Q1 question is, will these values will return in ascending order respective to the cluster IDs? So cluster 0 represents 0.2, cluster 1 represents 0.4, and cluster 2 represents 0.6? In Q2 question actually I don't want to know the physical location of the centroids, but only the KMeans, the values, like 0.2, 0.4 and 0.6 in this example. How to get those?

Dear Professor You made my day.excellent post.I love it.

Discussion

Earth Engine Tutorial #31: Machine Learning with Earth Engine - Unsupervised Classification

Responses(4)

Recent in Forum

Search Hashnode

Earth Engine Tutorial #31: Machine Learning with Earth Engine - Unsupervised Classification

Responses(4)

Recent in Forum