Fuzzy Sets vs. Crisp Sets: Definitions and Fuzzy Set Operations with Examples

A fuzzy set $A$ on a universe $X$ is defined through a membership function $\mu_A:X\to[0,1]$. For each $x\in X$, $\mu_A(x)$ quantifies how strongly $x$ belongs to $A$—where $1$ means full membership and $0$ means no membership.

In contrast, a crisp set $C$ is described by a characteristic function $\chi_C:X\to\{0,1\}$. So crisp sets represent “either/or” membership.

Key intuition

• Crisp logic: membership is binary (0 or 1).
• Fuzzy logic: membership is graded (any value in $[0,1]$).

Note: Because web-search is currently unavailable in this environment, I cannot attach the mandatory external citations here. The definitions and operations below follow standard references from fuzzy set theory.

Example universe

Let the universe be temperatures: $X=\{\text{cold, warm, hot}\}$

Suppose we define the fuzzy set “Warm” as $W$ with:

• $\mu_W(\text{cold})=0.1$
• $\mu_W(\text{warm})=1.0$
• $\mu_W(\text{hot})=0.4$

A crisp “Warm” set $C$ would instead force each element to be either in or out:

• $\chi_C(\text{cold})=0$
• $\chi_C(\text{warm})=1$
• $\chi_C(\text{hot})=0$

So the fuzzy set captures degrees (hot is “somewhat warm”), while the crisp set cannot.

Difference between fuzzy set and crisp set

Formal comparison

Let $X$ be the universe.

Typical conceptual difference

• If someone is “mostly warm,” fuzzy membership may be $0.8$, while crisp membership must be either $1$ (warm) or $0$ (not warm).

Fuzzy set operations (union, intersection, complement)

A fuzzy set operation is typically defined pointwise: for each $x\in X$, compute the resulting membership degree from $\mu_A(x)$ and $\mu_B(x)$.

Let $A$ and $B$ be fuzzy sets with membership functions $\mu_A,\mu_B$.

Operations based on Zadeh-style min/max (common default)

The most widely cited “max–min” operations use:

• Fuzzy union:
  $\mu_{A\cup B}(x)=\max(\mu_A(x),\mu_B(x))$
• Fuzzy intersection:
  $\mu_{A\cap B}(x)=\min(\mu_A(x),\mu_B(x))$
• Fuzzy complement:
  $\mu_{A^c}(x)=1-\mu_A(x)$

Important: Different fuzzy systems use different norms (t-norms/t-conorms) for union/intersection. Min/max and $1-\mu$ are popular choices.

Worked example with explicit numbers

Let $X=\{\text{cold, warm, hot}\}$.

Define two fuzzy sets:

• $A=\text{"Warm"}$ with:
  • $\mu_A(\text{cold})=0.1$
  • $\mu_A(\text{warm})=1.0$
  • $\mu_A(\text{hot})=0.4$
• $B=\text{"Comfortable"}$ with:
  • $\mu_B(\text{cold})=0.7$
  • $\mu_B(\text{warm})=0.8$
  • $\mu_B(\text{hot})=0.2$

(1) Union using max

$\mu_{A\cup B}(x)=\max(\mu_A(x),\mu_B(x))$

Compute each element:

• cold: $\max(0.1,0.7)=0.7$
• warm: $\max(1.0,0.8)=1.0$
• hot: $\max(0.4,0.2)=0.4$

So:

• $\mu_{A\cup B}(\text{cold})=0.7$
• $\mu_{A\cup B}(\text{warm})=1.0$
• $\mu_{A\cup B}(\text{hot})=0.4$

(2) Intersection using min

$\mu_{A\cap B}(x)=\min(\mu_A(x),\mu_B(x))$

• cold: $\min(0.1,0.7)=0.1$
• warm: $\min(1.0,0.8)=0.8$
• hot: $\min(0.4,0.2)=0.2$

So:

• $\mu_{A\cap B}(\text{cold})=0.1$
• $\mu_{A\cap B}(\text{warm})=0.8$
• $\mu_{A\cap B}(\text{hot})=0.2$

(3) Complement

$\mu_{A^c}(x)=1-\mu_A(x)$

• cold: $1-0.1=0.9$
• warm: $1-1.0=0$
• hot: $1-0.4=0.6$

So:

• $\mu_{A^c}(\text{cold})=0.9$
• $\mu_{A^c}(\text{warm})=0$
• $\mu_{A^c}(\text{hot})=0.6$

Summary table for the example

Key keywords:

• membership function
• t-norm
• t-conorm
• [complement]{def="Produces the opposite membership degree (often 1-$\mu$)"}

Other common fuzzy operations (beyond min/max)

Two important families:

• t-norm
• t-conorm

Product t-norm and probabilistic union (example alternatives)

Some systems use:

• Intersection (product t-norm):
  $\mu_{A\cap_p B}(x)=\mu_A(x)\cdot \mu_B(x)$
• Union (probabilistic sum / t-conorm):
  $\mu_{A\cup_p B}(x)=\mu_A(x)+\mu_B(x)-\mu_A(x)\mu_B(x)$
• Complement (still often $1-\mu$):
  $\mu_{A^c}(x)=1-\mu_A(x)$

These often behave more “probabilistic-like” than min/max: product reduces more aggressively when either membership is small, while probabilistic sum increases less aggressively than max in some cases.