Mastering Derivative Rules: Product, Quotient, and Chain

1. The Product Rule: Differentiating Multiplied Functions

Let's start with a question: if you have a function $h(x)=f(x)\cdot g(x)$, is the derivative simply $h^{\prime }(x)=f^{\prime }(x)\cdot g^{\prime }(x)$? It's a tempting idea, but it's incorrect. Let's see why.

As you saw, a more nuanced approach is required. The total change in the product depends on how each part contributes to the whole. A fantastic way to visualize this is to think of the function $h(x)=f(x)\cdot g(x)$ as the area of a rectangle with sides $f(x)$ and $g(x)$.

The intuition is that the total change is the sum of two contributions:

1. The change from $f$ (which is $df$) multiplied by the current size of $g$.
2. The change from $g$ (which is $dg$) multiplied by the current size of $f$.

This leads directly to the formal Product Rule.

In Leibniz notation, if $h(x)=f(x)g(x)$, then:

Let's see this in action.

2. The Quotient Rule: Differentiating Divided Functions

Just as with products, we can't simply differentiate the numerator and denominator separately. We need a specific rule for quotients of the form $h(x)=\frac{f(x)}{g(x)}$.

The formula is similar in spirit to the product rule but with a crucial subtraction and a new term in the denominator.

If $h(x)=\frac{f(x)}{g(x)}$, the Quotient Rule is:

A common mnemonic to remember this is: "Low D-high minus High D-low, square the bottom and away we go!", where "Low" is $g(x)$, "High" is $f(x)$, and "D" means "the derivative of".

Let's watch this rule being applied.

3. The Chain Rule: Differentiating Composite Functions

The Product and Quotient rules handle functions that are multiplied or divided. But what about functions that are nested inside one another, like $h(x)=\sin (x^2)$? This is called a composite function. To differentiate these, we need the Chain Rule.

The core idea of the Chain Rule is to differentiate from the "outside in". You take the derivative of the outermost function, leaving the inner function untouched, and then you multiply by the derivative of the inner function.

Let's build our intuition for this first.

This "outside-in" process is formalized in the Chain Rule.

If $h(x)=f(g(x))$, then:

Let's see this put to work with some clear, step-by-step examples.

The Chain Rule can be applied multiple times for deeply nested functions, like peeling an onion one layer at a time.

4. Combining the Rules

True mastery comes from being able to combine these rules to tackle complex functions. You might encounter a product where one of the factors is a composite function, or a quotient where the numerator requires the chain rule. The key is to identify the main structure first (product, quotient, or chain) and then apply the other rules as needed.

Let's look at an example where the Product Rule and Chain Rule are used together.

This systematic, hierarchical approach is exactly the kind of multi-step calculation you aspire to master. Identify the outermost operation, apply the corresponding rule, and then solve the smaller derivative problems that arise, one step at a time.

Conclusion

In this lesson, you've added the three most important differentiation rules to your arsenal. These rules allow you to move from simple functions to the complex, interconnected structures that define real-world models.

Key Takeaways:

• The Product Rule: Used for functions of the form $f(x)\cdot g(x)$. The derivative is $f^{\prime }g+f{g}^{\prime }$.
• The Quotient Rule: Used for functions of the form $\frac{f(x)}{g(x)}$. The derivative is $\frac{f^{\prime }g-f{g}^{\prime }}{g^2}$.
• The Chain Rule: Used for composite functions of the form $f(g(x))$. The derivative is $f^{\prime }(g(x))\cdot g^{\prime }(x)$, working from the "outside in".
• Complex derivatives are solved by identifying the main structure first and applying rules hierarchically.

Preview of the Next Lesson:
So far, all our functions have had a single input, $x$. However, the loss functions in machine learning depend on millions of parameters (weights and biases). To understand how to optimize these, we need to extend our calculus to functions of multiple variables. In the next lesson, you will learn to compute partial derivatives of multivariable functions. You will find that the rules you mastered today are the direct building blocks for this more advanced topic.