Model answer

Picture each class as a fuzzy cloud floating in color-space. Max Likelihood asks: “Which cloud is this pixel most likely from?”

Crucially, it accounts for each cloud’s shape — not just its center. A tight class is treated differently from a sprawling one.

The catch: it assumes the clouds are roughly Gaussian blobs. If your classes are weird-shaped (banana-shaped, donut-shaped), ML’s accuracy drops.

It’s the most popular supervised classifier — but only when the data fits the assumption.

🔬 Show the science / technical version

• Assumes each class’s pixels follow a multivariate Gaussian distribution.
• For each pixel, compute probability of membership in every class (from class mean + covariance).
• Assign pixel to the class with the highest probability.
• Strength: statistically principled, handles correlated bands.
• Weakness: needs enough training samples to estimate covariance; struggles if classes aren’t Gaussian.