Numerical methods for linear least squares are a way of finding the best solution to a problem when you know two things about it.
Say you're ordering pizzas for a party. You want to figure out how many of each kind to order so that everyone gets their favorite. But there are too many kinds to count, so you want to find the best way to do it - the way that makes the most people happy.
The first thing you know is that the number of each kind of pizza you order needs to add up to the total number of guests. This is like having an equation - we can say the total number of pizzas equals the number of guests.
The second thing you know is that each person should get their favorite pizza. So, you ask each guest what kind they want, so you can make sure everyone gets exactly what they want.
Numerical methods for linear least squares are a way to use those two pieces of information to figure out how many of each kind of pizza to order. It's like a guessing game - you start with an estimate of how many you need of each kind, and then you modify your answer until you come up with an answer that makes the most people happy. To do this, you use a combination of math and numbers to test out each possible solution. Once you find the answer that makes the most people happy and still adds up to the total number of guests, you have your answer - the best way to order the pizzas for the party.