What is a Derivative?

1 Rate of Change , The Core Idea

You already have the intuition for this, you just might not have called it a derivative. Speed is how fast your position changes. Acceleration is how fast your speed changes. The steepness of a hill is how fast elevation changes with horizontal distance. All of those are rates of change, which is all a derivative is. , 60 mph means your position changes by 60 miles per hour. Acceleration is rate of change of speed. The steepness of a hill is the rate of change of elevation for horizontal distance.

Average rate of change between two points is easy: (change in output) / (change in input) = Δy/Δx. For a car traveling 120 miles in 2 hours, average speed = 120/2 = 60 mph.

Average speed over an hour is easy: divide distance by time. But your speed at exactly one specific instant is a different question. At 1:32 PM precisely, not over a minute, not over a second. That instantaneous rate of change is what a derivative computes. , not over an hour, not over a minute, but at a single instant? That instantaneous rate of change is what the derivative computes.

2 Derivative as Slope

On a graph, pick two points on a curve and draw a line between them. That line has a slope, which is the average rate of change between those two points. Now slide one point toward the other. The line rotates as you do. When the two points get infinitely close, that rotating line becomes the tangent: a line that just grazes the curve at one point. The slope of that tangent is the derivative., the secant line rotates and approaches the tangent line , a line that just touches the curve at one point.

The derivative at a point is the slope of the tangent line at that point. For a straight line y = mx + b, the slope is constant everywhere , the derivative is m at every point. For a curve like y = x², the slope changes at every point, and the derivative tells you exactly what that slope is.

3 The Formal Definition

f'(x) = lim(h→0) [f(x+h) − f(x)] / h

This is the difference quotient taken to its limit. [f(x+h) − f(x)] / h is the slope of the secant line between x and x+h. As h→0, this approaches the slope of the tangent at x.

This means: at any point x on the parabola y=x², the slope of the tangent line is 2x. At x=3, the slope is 6. At x=−1, the slope is −2. At x=0 (the vertex), the slope is 0.

4 Derivative Notation

Several notations exist for the derivative, all meaning the same thing:

5 Real-World Meaning

Physics: velocity is the derivative of position for time. Acceleration is the derivative of velocity. Force equals mass times acceleration , Newton's second law is fundamentally about derivatives.

Economics: marginal cost is the derivative of total cost for quantity , the cost of producing one more unit. Marginal revenue is the derivative of revenue. Profit is maximized where marginal cost = marginal revenue, which is a derivative condition.

Medicine: pharmacokinetics models how drug concentration in blood changes over time using derivatives. The rate at which a drug clears the body is described by a differential equation , an equation involving derivatives.

In machine learning, training a neural network means minimizing a loss function, which is done by computing gradients, which are derivatives. Every AI model that exists was trained using derivatives to figure out which direction to adjust its parameters. The derivative tells the optimization algorithm which way is downhill. , moving in the direction of the negative derivative (gradient) to find the minimum. Every AI model you've ever used was trained with derivatives.

Sources & Further Reading

The explanations on this page draw on the following established sources. We link to primary and secondary sources so you can verify claims and go deeper on any topic.