optimization

Optimization is the art of finding the best — the most profit, the least cost, the largest volume, the shortest path. A farmer with 100 metres of fencing wants the biggest rectangular paddock; a soup company wants the can that holds a fixed volume using the least metal. In every case there is some quantity you want to push as high or as low as it will go, and calculus turns that wish into a calculation.

The workflow is reliable once you see it. First, write the quantity you care about as a function — the objective. If it depends on more than one variable, use a constraint (like the fixed perimeter or fixed volume) to eliminate variables until the objective depends on just one. Then find the critical points by setting the derivative to zero, and finally compare the candidate values, including the endpoints of the allowed range, to see which truly gives the maximum or minimum. The phrase to remember is: function -> constraint -> critical points -> check endpoints.

The step people skip is the last one, and it bites. A critical point is only a candidate; the true best value might sit at an endpoint of the feasible region, or the second-derivative test might reveal a critical point is the wrong kind of extreme. Always confirm which critical point wins and check the boundaries — the derivative finds flat spots, but it does not by itself know whether they are peaks, valleys, or the edges of what is allowed.