A theoretical visualization of the bias-variance tradeoff showing how total prediction error decomposes into bias squared, variance, and irreducible noise as a function of model complexity. The plot displays multiple curves: bias squared (decreasing with complexity), variance (increasing with complexity), irreducible error (constant), and total error (U-shaped). This is one of the most fundamental conceptual plots in machine learning for understanding model selection, overfitting, and underfitting.

// anyplot.ai
// curve-bias-variance-tradeoff: Bias-Variance Tradeoff Curve
// Library: echarts 6.1.0 | JavaScript 22.23.2
// Quality: 89/100 | Created: 2026-08-24
const t = window.ANYPLOT_TOKENS;
// --- Data (in-memory, deterministic) ----------------------------------------
// model_complexity spans 1..20 (e.g. polynomial degree / tree depth)
const N = 60;
const complexity = Array.from({ length: N }, (_, i) => 1 + (i * 19) / (N - 1));
const biasSquared = complexity.map((c) => 4 / (1 + 0.55 * c) + 0.15);
const variance = complexity.map((c) => 0.012 * c * c + 0.03);
const irreducible = complexity.map(() => 0.35);
const totalError = complexity.map((_, i) => biasSquared[i] + variance[i] + irreducible[i]);
// Find the optimal complexity (minimum total error)
let optIdx = 0;
for (let i = 1; i < N; i++) {
if (totalError[i] < totalError[optIdx]) optIdx = i;
}
const optComplexity = complexity[optIdx];
const optError = totalError[optIdx];
// --- Init ---------------------------------------------------------------
const chart = echarts.init(document.getElementById("container"));
// --- Option ---------------------------------------------------------------
chart.setOption({
animation: false,
color: [t.palette[0], t.palette[2], t.palette[4], t.palette[6]],
backgroundColor: "transparent",
title: {
text: "curve-bias-variance-tradeoff · javascript · echarts · anyplot.ai",
left: "center",
top: 12,
textStyle: { color: t.ink, fontSize: 20, fontWeight: 500 },
},
grid: { left: 90, right: 170, top: 150, bottom: 90 },
legend: {
top: 60,
textStyle: { color: t.ink, fontSize: 15 },
itemWidth: 26,
itemHeight: 3,
},
xAxis: {
type: "value",
name: "Model Complexity",
nameLocation: "middle",
nameGap: 42,
nameTextStyle: { color: t.ink, fontSize: 16 },
min: 1,
max: 20,
axisLabel: { color: t.inkSoft, fontSize: 14 },
axisLine: { lineStyle: { color: t.inkSoft } },
axisTick: { show: false },
splitLine: { show: false },
},
yAxis: {
type: "value",
name: "Prediction Error",
nameLocation: "middle",
nameGap: 60,
nameTextStyle: { color: t.ink, fontSize: 16 },
min: 0,
axisLabel: { color: t.inkSoft, fontSize: 14 },
axisLine: { show: true, lineStyle: { color: t.inkSoft } },
axisTick: { show: false },
splitLine: { lineStyle: { color: t.grid } },
},
series: [
{
name: "Bias²",
type: "line",
data: complexity.map((c, i) => [c, biasSquared[i]]),
showSymbol: false,
lineStyle: { width: 3.5, color: t.palette[0] },
endLabel: {
show: true,
formatter: "{a}",
color: t.palette[0],
fontSize: 14,
fontWeight: 500,
distance: 10,
offset: [0, -16],
},
z: 3,
markArea: {
silent: true,
itemStyle: { color: t.palette[0], opacity: 0.1 },
label: { position: "insideTop", color: t.inkSoft, fontSize: 15 },
data: [[{ xAxis: 1, name: "Underfitting zone" }, { xAxis: optComplexity }]],
},
},
{
name: "Variance",
type: "line",
data: complexity.map((c, i) => [c, variance[i]]),
showSymbol: false,
lineStyle: { width: 3.5, color: t.palette[2] },
endLabel: {
show: true,
formatter: "{a}",
color: t.palette[2],
fontSize: 14,
fontWeight: 500,
distance: 10,
},
z: 3,
markArea: {
silent: true,
itemStyle: { color: t.palette[2], opacity: 0.1 },
label: { position: "insideTop", color: t.inkSoft, fontSize: 15 },
data: [[{ xAxis: optComplexity, name: "Overfitting zone" }, { xAxis: 20 }]],
},
},
{
name: "Irreducible Error",
type: "line",
data: complexity.map((c, i) => [c, irreducible[i]]),
showSymbol: false,
lineStyle: { width: 2.5, color: t.palette[6], type: "dashed" },
endLabel: {
show: true,
formatter: "{a}",
color: t.palette[6],
fontSize: 14,
fontWeight: 500,
distance: 10,
offset: [0, 16],
},
z: 2,
},
{
name: "Total Error",
type: "line",
data: complexity.map((c, i) => [c, totalError[i]]),
showSymbol: false,
lineStyle: { width: 4, color: t.palette[4] },
endLabel: {
show: true,
formatter: "{a}",
color: t.palette[4],
fontSize: 14,
fontWeight: 600,
distance: 10,
},
z: 4,
markLine: {
silent: true,
symbol: "none",
lineStyle: { color: t.ink, type: "dashed", width: 1.5 },
label: {
formatter: "Optimal complexity",
color: t.ink,
fontSize: 14,
position: "insideEndTop",
},
data: [{ xAxis: optComplexity }],
},
markPoint: {
silent: true,
symbol: "circle",
symbolSize: 14,
itemStyle: { color: t.palette[4], borderColor: t.pageBg, borderWidth: 2 },
label: { show: false },
data: [{ coord: [optComplexity, optError] }],
},
},
],
graphic: [
{
type: "text",
left: 90,
top: 110,
style: {
text: "Total Error = Bias² + Variance + Irreducible Error",
fill: t.inkSoft,
fontSize: 15,
fontStyle: "italic",
},
},
],
});
Part of Bias-Variance Tradeoff Curve on anyplot.ai.