What type of learning is K means?
Unsupervised learning where the data has no labels.
What is the goal of K means clustering?
To group data points into k clusters based on similarity.
What does k represent in K means?
The number of clusters you want the algorithm to find.
How does K means start the clustering process?
It randomly selects k data points as initial cluster centres.
What is a centroid?
The mean of the data points in a cluster.
How does K means assign points to clusters?
It assigns each point to the cluster whose centroid it is closest to.
When does the K means algorithm stop?
When the centroids no longer change between iterations.
Why do points move between clusters?
Because the centroids get recalculated and new distances change the assignment.
Can K means predict cluster membership for new data?
Yes, by checking which centroid the new point is closest to.
Why should K means forecasts not be fully relied on?
Because new data changes the cluster means, which affects predictions.
What is the elbow method?
A way to choose k by looking at where the WSS curve starts to flatten.
What is WSS?
Within sum of squares, a measure of how tight the clusters are.
Give one workplace use for clustering in FPA.
Grouping customers or spend patterns to find behaviour trends.
Why is clustering useful with multi dimensional data?
Because it groups points across many different measures at once.
Why did you use Python instead of R?
IT could not approve R, and Python is already available and works for clustering.
