#datawithSeinfeld 2 of 10: The Red Dot 🔴

George Costanza, ever the bargain hunter, was thrilled to see the discount on the cashmere sweater - despite the tiny red dot. He buys the sweater and then gifts it to Elaine (and then Evie, the cleaning lady).

Unfortunately for George, Elaine (and Kramer, and Evie) noticed the red dot immediately and no longer wanted the gift.

Object Detection is a popular tool in machine learning - where algorithms are trained to identify objects in images - like red dots. The machines aren't perfect, and they sometimes "see" things that aren't there, or miss things that are. Two key metrics are recall and precision:

1️⃣ Recall is the ability of a model to find all relevant cases. In George's case, this means being able to spot every red dot.

2️⃣ Precision is the ability of the model to identify ONLY the relevant objects. For George, this would mean only spotting ACTUAL red dots.

Well would you believe it, poor George had a 103 fever that day and was seeing red dots everywhere! His recall was high...but his precision was terrible!

In analytics there's a trade-off between recall and precision. It's a delicate balance, and deciding where to set a threshold is an important part of any object detection tool.

Next time you're picking out a cashmere sweater, make sure your object detector is tuned appropriately!