Little Bits of Artificial Intelligence

A. Our Approach to Unsupervised Learning (like K-Means):

1.We randomly pick some starting points (called centroids — just like guesses of where the centers of groups might be).

2. Then, for each data point, we:

• Check how close it is to each of these centroids
• Group it with the nearest one

3. Once all the points are grouped:

• We recalculate the real center (centroid) of each group (based on where the points are now)
• Then we repeat the process:
• Check distances again
• Reassign groups if needed
• Update centroids

4. After a few rounds, the groups settle down, and we have our final clusters!

We Can Think of It Like This:

It’s like playing a game of hot-and-cold:
We guess a center, the points gather around the closest center,we adjust the centers to better fit the groups, and repeat until everything’s stable.

B. Real-Life Example:

Imagine we’re organizing:

• Kids into teams based on height and weight
• Or toys based on shape and size

We don’t know the teams ahead of time, but we group them based on who looks or feels similar — and update the group centers as we go.
We’re doing logical grouping based on closeness to some starting points, and we keep adjusting to make better groups.