Detailed slide-by-slide text content extracted from this presentation.
Slide 01
Fractal Geometry
- The Geometry of Nature
- Infinite complexity from simple rules -- the mathematics of roughness, self-similarity, and fractional dimensions
- Self-Similarity
- Mandelbrot Set
- Iterated Functions
- Chaos Theory
- Fractional Dimensions
- 1 / 23
Slide 02
What Are Fractals?
- A fractal is a geometric object that displays self-similarity at different scales -- zooming in reveals structures that echo the whole. Unlike the smooth shapes of Euclidean geometry (lines, circles, spheres), fractals are rough, fragmented, and infinitely detailed.
- Key Properties
- Self-similarity: Parts resemble the whole (exact, quasi, or statistical)
- Fractional dimension: Dimension is not a whole number (e.g., 1.26, 2.58)
- Infinite detail: New structure appears at every magnification
- Simple rules, complex output: Generated by iteration of simple formulas
- Euclidean vs. Fractal
- Euclidean: smooth, finite, integer dimensions
- Fractal: rough, infinite detail, fractional dimensions
- Euclidean: describes human-made objects (buildings, roads)
- Fractal: describes natural objects (mountains, coastlines, trees)
- "Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line." -- Benoit Mandelbrot, The Fractal Geometry of Nature (1982)
- 2 / 23
Slide 03
Benoit Mandelbrot
- Born in Warsaw in 1924, Benoit B. Mandelbrot was a Polish-French-American mathematician who coined the word "fractal" in 1975 (from Latin fractus, meaning "broken") and almost single-handedly created fractal geometry as a field.
- Life and Career
- Family fled Nazi-occupied France; largely self-taught during the war years
- PhD from the University of Paris (1952) under Paul Levy
- Spent 35 years at IBM's Thomas J. Watson Research Center (1958-1993)
- Sterling Professor of Mathematical Sciences at Yale (1999-2010)
- Published The Fractal Geometry of Nature in 1982, making fractals accessible to scientists and the public
- Died October 14, 2010, aged 85
- "I conceived and developed a new geometry of nature and implemented its use in a number of diverse fields. It describes many of the irregular and fragmented patterns around us." -- Benoit Mandelbrot
- Mandelbrot never won the Fields Medal but received the Wolf Prize (1993), the Japan Prize (2003), and numerous other honors. His work influenced mathematics, physics, biology, economics, art, and computer science.
- 3 / 23
Slide 04
Before Mandelbrot: Early Fractals
- Several "pathological" mathematical objects discovered in the 19th and early 20th centuries were later recognized as fractals. At the time, they were dismissed as "monsters" -- curiosities with no practical value.
- 1872
- Weierstrass function: A continuous function that is nowhere differentiable -- infinitely jagged at every point. Shocked mathematicians who assumed continuity implied smoothness.
- 1883
- Cantor set: Georg Cantor removes the middle third of a line segment, repeatedly. The result has zero length but uncountably many points. Dimension: log(2)/log(3) = 0.631.
- 1904
- Koch snowflake: Helge von Koch constructs a curve of infinite length enclosing finite area. Dimension: log(4)/log(3) = 1.262.
- 1915
- Sierpinski triangle: Waclaw Sierpinski creates a triangle with zero area by recursively removing central triangles. Dimension: log(3)/log(2) = 1.585.
- 1918
- Hausdorff dimension: Felix Hausdorff defines a rigorous notion of fractional dimension, providing the mathematical foundation fractals would later need.
- 4 / 23
Slide 05
Fractal Dimension
- In Euclidean geometry, a line is 1-dimensional, a plane is 2-dimensional, and a solid is 3-dimensional. Fractals break this rule: their dimension is typically a non-integer, capturing how completely they fill space.
- Self-Similarity Dimension
- D = log(N) / log(S)
- N = number of self-similar pieces, S = scaling factor
- FractalCopies (N)Scale (S)Dimension (D)
- Cantor set230.631
- Koch curve431.262
- Sierpinski triangle321.585
- Sierpinski carpet831.893
- Menger sponge2032.727
- The box-counting dimension generalizes this: cover the fractal with boxes of side length ε and count how the number of boxes N(ε) scales as ε shrinks. D = -lim(log N(ε) / log ε).
- 5 / 23
Slide 06
The Mandelbrot Set
- The most famous fractal in mathematics. Discovered/visualized by Mandelbrot in 1980 using IBM computers, it is defined by a deceptively simple formula in the complex plane.
- Defining Iteration
- zn+1 = zn2 + c   where z0 = 0
- A complex number c belongs to the Mandelbrot set if this iteration does not diverge to infinity. The boundary of this set is infinitely complex -- zoom in anywhere and you find miniature copies of the whole set, embedded in swirling filaments of extraordinary beauty.
- Key Facts
- The boundary has a Hausdorff dimension of exactly 2 (proven by Mitsuhiro Shishikura, 1991)
- The set is connected (proven by Adrien Douady and John Hubbard, 1982)
- Whether the Mandelbrot set is locally connected remains an open problem
- The area of the Mandelbrot set is approximately 1.5065 square units
- Zooming in reveals infinite variety: spirals, seahorses, lightning bolts, satellite copies
- 6 / 23
Slide 07
Julia Sets
- Named after French mathematician Gaston Julia (1893-1978), who studied these objects in 1918 -- decades before computers could visualize them. Each point c in the complex plane defines a unique Julia set.
- Julia Set Iteration
- zn+1 = zn2 + c   (c is fixed; z0 varies)
- Mandelbrot-Julia Correspondence
- If c is inside the Mandelbrot set, its Julia set is connected (one piece)
- If c is outside the Mandelbrot set, its Julia set is a Cantor dust (infinitely many disconnected pieces)
- The Mandelbrot set is, in a sense, a "catalog" of all possible Julia sets
- Famous Julia Sets
- c = -0.7 + 0.27i
- "Dragon" Julia set with spiral arms and baroque detail
- c = -0.8 + 0.156i
- "Dendrite" Julia set -- tree-like branching structure
- c = 0.285 + 0.01i
- "Siegel disk" Julia set with concentric ring structures
- 7 / 23
Slide 08
Iterated Function Systems
- An Iterated Function System (IFS) is a collection of contraction mappings that, when applied repeatedly, produce a fractal as their "attractor." Formalized by Michael Barnsley in the 1980s.
- How IFS Works
- Define a set of affine transformations (scale, rotate, translate)
- Start with any initial point
- Randomly choose a transformation and apply it
- Plot the result and repeat thousands of times
- The fractal emerges as the limit of this process
- Barnsley's Fern
- Just four affine transformations generate a strikingly realistic fern leaf. The stem, left frond, right frond, and overall shape are each produced by a separate mapping. Probabilities: stem 1%, left leaflet 7%, right leaflet 7%, successive leaflet 85%.
- Fractal Compression
- Barnsley's student Arnaud Jacquin developed fractal image compression (1989), which encodes images as IFS codes. This achieved high compression ratios but was computationally expensive. It influenced later work on wavelet compression.
- 8 / 23
Slide 09
L-Systems: Fractal Grammars
- Lindenmayer systems (L-systems), invented by biologist Aristid Lindenmayer in 1968, use string-rewriting rules to generate fractal structures. Originally designed to model plant growth.
- How They Work
- Alphabet: A set of symbols (e.g., F, +, -, [, ])
- Axiom: The starting string (e.g., "F")
- Production rules: Replacement rules applied in parallel (e.g., F → F+F-F-F+F)
- Interpretation: Symbols are converted to drawing commands (turtle graphics)
- Examples
- Koch Curve
- Rule: F → F+F-F+F
- Angle: 60 degrees
- Iterations: 4-6 produce a recognizable snowflake
- Dragon Curve
- Rule: F → F+G, G → F-G
- Angle: 90 degrees
- Unfolds a paper-folding sequence into a space-filling curve
- Fractal Plant
- Rule: X → F+[[X]-X]-F[-FX]+X
- Branching with [/] for push/pop state. Produces realistic botanical forms.
- Pixar, Weta Digital, and other VFX studios use L-system variants to generate the forests, grass, and vegetation seen in films.
- 9 / 23
Slide 10
Fractals in Nature
- Mandelbrot's key insight was that fractal geometry describes the shapes of nature far better than Euclidean geometry. Natural fractals are not perfectly self-similar but exhibit statistical self-similarity.
- Biological Fractals
- Lungs: 23 levels of bronchial branching pack ~70 m² of surface into your chest
- Blood vessels: Fractal branching network totals ~100,000 km in an adult
- Neurons: Dendritic trees are fractal; their dimension correlates with function
- Romanesco broccoli: Near-perfect logarithmic spiral self-similarity
- Fern fronds: Each leaflet resembles the whole frond
- Geological Fractals
- Coastlines: Mandelbrot's paper "How Long Is the Coast of Britain?" (1967) launched the field
- Mountains: Fractal dimension ~2.2-2.5 for natural terrain
- Rivers: Drainage networks follow Horton's law, a fractal scaling relation
- Clouds: Boundary has dimension ~1.35
- Lightning: Fractal branching pattern, D ~ 1.7
- "Why is geometry often described as 'cold' and 'dry'? One reason lies in its inability to describe the shape of a cloud, a mountain, a coastline, or a tree." -- Mandelbrot, 1982
- 10 / 23
Slide 11
The Coastline Paradox
- In 1967, Mandelbrot published "How Long Is the Coast of Britain? Statistical Self-Similarity and Fractional Dimension" -- the paper that launched fractal geometry.
- The Problem
- The measured length of a coastline depends on the length of the ruler you use. A 100 km ruler gives one answer; a 1 km ruler gives a much longer answer because it follows more indentations; a 1 m ruler gives an even longer answer. As the ruler shrinks toward zero, the measured length diverges to infinity.
- 2,800 kmBritain's coast (100 km ruler)
- 3,400 km50 km ruler
- ~17,800 kmOrdnance Survey (detailed)
- Mandelbrot's Answer
- A coastline has no well-defined length. Instead, its fractal dimension quantifies its roughness. Norway (deeply fjorded) has D ~ 1.52; Britain has D ~ 1.25; South Africa (smooth) has D ~ 1.02. The higher the dimension, the more the coastline "fills" the plane.
- 11 / 23
Slide 12
Fractals and Chaos
- Fractal geometry and chaos theory are deeply intertwined. Chaotic dynamical systems often produce fractal structures in their phase space -- and vice versa.
- Strange Attractors
- In a chaotic system, trajectories converge to a "strange attractor" -- a fractal set in phase space. The system is deterministic but unpredictable over long times.
- Lorenz Attractor (1963)
- Edward Lorenz discovered chaos while modeling weather. Three simple differential equations produce a butterfly-shaped attractor with dimension ~2.06. This is the origin of the "butterfly effect."
- Henon Map (1976)
- A 2D discrete map: xn+1 = 1 - axn² + yn, yn+1 = bxn. Classic parameters a=1.4, b=0.3 give an attractor with D ~ 1.26. Zoom in to see fractal layering.
- The Logistic Map
- xn+1 = rxn(1 - xn). As the parameter r increases from 1 to 4, the system transitions from stability through period-doubling cascades to chaos. The bifurcation diagram is a fractal with universal scaling ratios discovered by Mitchell Feigenbaum (1975).
- 12 / 23
Slide 13
Space-Filling Curves
- A space-filling curve is a continuous curve that passes through every point of a 2D region. Though 1-dimensional by definition, it has Hausdorff dimension 2. These objects astounded 19th-century mathematicians.
- Peano Curve (1890)
- Giuseppe Peano constructed the first space-filling curve, proving that a 1D line can map surjectively onto a 2D square. This challenged the very notion of dimension.
- Hilbert Curve (1891)
- David Hilbert's variant is simpler to construct and widely used in computer science for spatial indexing. Google's S2 geometry library uses Hilbert curves to index the Earth's surface.
- Dragon Curve
- Discovered by John Heighway (1966). Unfold a strip of paper folded in half repeatedly -- the limit is a space-filling curve that tiles the plane.
- Z-Order (Morton) Curve
- Interleaves binary coordinates to map 2D/3D points to 1D. Used in database indexing, GPU memory access patterns, and octree data structures.
- Space-filling curves preserve locality: nearby points in 2D tend to be nearby on the curve. This property makes them valuable for cache-efficient algorithms and spatial database queries.
- 13 / 23
Slide 14
Fractals in Physics
- Fractals appear throughout physics, from the quantum scale to the cosmic web. They arise naturally in systems with scale-invariance -- where the same physics operates at every length scale.
- Percolation Theory
- At the critical threshold where a material transitions from non-conducting to conducting, the connected cluster forms a fractal with D ~ 1.896 (2D). Applied to oil recovery, forest fire spread, and epidemiology.
- Diffusion-Limited Aggregation
- DLA (Witten & Sander, 1981) simulates particles randomly walking until they stick to a growing cluster. Produces dendritic fractals resembling snowflakes, mineral deposits, and electrochemical growth. D ~ 1.71 in 2D.
- Turbulence
- Richardson's 1922 poem: "Big whirls have little whirls that feed on their velocity, and little whirls have lesser whirls and so on to viscosity." Turbulent flows have fractal energy cascades.
- The Cosmic Web
- The large-scale structure of the universe -- filaments, walls, and voids of galaxy clusters -- displays fractal-like scaling up to ~300 Mpc, beyond which it becomes homogeneous.
- 14 / 23
Slide 15
Fractals in Medicine
- The human body is full of fractal structures -- and deviations from normal fractal patterns can signal disease.
- Diagnostic Applications
- Heart rate variability: A healthy heartbeat is fractal (not perfectly regular). Loss of fractal complexity in the RR interval predicts cardiac events. Goldberger et al. (2002) showed that both extreme regularity and extreme randomness are pathological.
- Cancer detection: Tumor vasculature has a different fractal dimension than normal tissue. Fractal analysis of mammograms can detect suspicious patterns. Studies have shown D ~ 1.7 for malignant masses vs. D ~ 1.3 for benign.
- Retinal imaging: The fractal dimension of retinal blood vessels decreases with age and in diabetes, hypertension, and glaucoma. Automated fractal analysis aids screening.
- Neuroscience: Fractal dimension of cortical folding (gyrification) correlates with intelligence and changes in schizophrenia and Alzheimer's disease.
- Bone structure: Trabecular (spongy) bone has fractal architecture. Osteoporosis reduces its fractal dimension, detectable on dental X-rays.
- 15 / 23
Slide 16
Fractals in Finance
- Mandelbrot's work on financial markets was controversial and prescient. He showed that market prices are not well described by standard models -- they are fractal.
- The Problem with Normal Distributions
- The standard Black-Scholes model assumes stock returns follow a normal distribution. But real markets have "fat tails" -- extreme events occur far more often than a bell curve predicts. The 2008 financial crisis was a "25-sigma event" under normal assumptions -- essentially impossible.
- Mandelbrot's Alternative
- Cotton price research (1963): Mandelbrot found that cotton prices followed a Levy stable distribution with infinite variance -- not the Gaussian
- Multifractal model of asset returns (1997): Prices have different fractal properties at different time scales -- they are multifractal, not monofractal
- Hurst exponent: H > 0.5 indicates persistent (trending) behavior; H
- "The financial markets are fractal in nature. The reason is that the markets are driven by human beings, and human beings are fractal in their behavior." -- Mandelbrot, The (Mis)Behavior of Markets (2004)
- 16 / 23
Slide 17
Fractals in Computer Graphics
- Fractal algorithms revolutionized computer-generated imagery by making it possible to create photorealistic natural scenery procedurally.
- Terrain Generation
- The midpoint displacement algorithm (diamond-square) and Perlin noise generate fractal landscapes used in games and films. Loren Carpenter's 1980 demo of fractal mountains led to his co-founding of Pixar.
- Fractal Flames
- Invented by Scott Draves (1992), fractal flames use nonlinear IFS with artistic coloring. The Electric Sheep screensaver and many music visualizers use this technique. Apophysis and JWildfire are popular renderers.
- Mandelbulb and 3D Fractals
- The Mandelbulb (2009) extends the Mandelbrot set to 3D using spherical coordinates. Ray-marched using signed distance functions, it produces stunning volumetric structures explored in real-time with GPU shaders.
- Procedural Textures
- Noise functions (Perlin, Simplex, Worley) create fractal-based textures for wood grain, marble, clouds, water, and fire. Used in virtually every 3D renderer since the 1980s.
- Films like Star Trek II: The Wrath of Khan (1982, Genesis sequence), Frozen (ice crystals), and Avatar (Pandora's forests) relied heavily on fractal generation techniques.
- 17 / 23
Slide 18
Fractal Art
- Fractals bridge mathematics and aesthetics. The infinite detail and emergent complexity of fractal images have inspired a thriving artistic movement since the 1980s.
- Pioneers and Movements
- The Mandelbrot set images (1980s): Among the first computer-generated images to achieve widespread cultural recognition. Posters sold millions of copies.
- Kerry Mitchell's Fractal Art Manifesto (1999): Argued that fractal art is a legitimate art form, not merely a mathematical curiosity
- Jackson Pollock: Analysis by physicist Richard Taylor (1999) showed that Pollock's drip paintings have fractal dimension D ~ 1.7, increasing over his career from D ~ 1.3 (1943) to D ~ 1.7 (1952)
- Tools and Software
- Ultra FractalProfessional fractal art tool
- Mandelbulb3D3D fractal explorer
- ShadertoyReal-time GPU fractals
- XaoSReal-time fractal zoom
- 18 / 23
Slide 19
Fractals in Engineering
- The self-similar, space-filling properties of fractals have practical engineering applications, particularly in antenna design and signal processing.
- Fractal Antennas
- Nathan Cohen (1988) discovered that antennas shaped as fractals (Koch, Sierpinski, Minkowski) can operate efficiently at multiple frequencies simultaneously and in a smaller form factor. Your smartphone likely contains a fractal antenna. Key advantage: a single antenna covers GSM, WiFi, Bluetooth, and GPS bands.
- Heat Exchangers
- Fractal branching patterns (inspired by lungs and blood vessels) maximize surface area for heat transfer while minimizing volume. Used in HVAC systems and microelectronics cooling.
- Signal and Image Processing
- Wavelet transforms, closely related to fractal analysis, are used in JPEG 2000 compression, fingerprint storage (FBI), and seismographic data analysis. Wavelets decompose signals at multiple scales -- a fundamentally fractal operation.
- Urban Planning
- Cities exhibit fractal growth patterns. Batty and Longley (1994) showed that urban boundaries have D ~ 1.3-1.7. Fractal analysis helps planners model sprawl, optimize infrastructure, and design transport networks.
- 19 / 23
Slide 20
Multifractals
- A simple fractal has a single fractal dimension throughout. A multifractal has a spectrum of dimensions -- different regions scale differently. Most real-world fractals are multifractal.
- The Multifractal Spectrum
- The singularity spectrum f(α) characterizes how the local scaling exponent α is distributed across the fractal. A monofractal has a single peak; a multifractal has a broad, parabolic spectrum.
- Generalized Dimensions Dq
- Dq = (1/(q-1)) · lim log(Σ piq) / log(ε)
- Applications
- Turbulence: Energy dissipation in turbulent flows is multifractal (Kolmogorov refined his 1941 theory in 1962 to account for this)
- Financial markets: Mandelbrot's MMAR model uses multifractal measures to capture volatility clustering
- Geophysics: Earthquake magnitude distributions and mineral ore deposits exhibit multifractal scaling
- Medical imaging: Multifractal analysis of MRI and CT scans can distinguish healthy from diseased tissue
- 20 / 23
Slide 21
Fractals, Complexity, and Emergence
- Fractals exemplify one of the deepest themes in modern science: how complex structures emerge from simple rules through iteration and feedback.
- Cellular Automata
- Stephen Wolfram's Rule 90 and Rule 150 produce Sierpinski triangles from 1D binary automata. Conway's Game of Life generates complex patterns from four simple rules. Wolfram argued in A New Kind of Science (2002) that such systems underlie all of nature.
- Self-Organized Criticality
- Per Bak's sandpile model (1987) showed that many systems naturally evolve to a critical state where avalanches follow power-law (fractal) distributions. Applications: earthquakes, forest fires, neural activity, stock market crashes.
- The Edge of Chaos
- Complex systems often exhibit the richest behavior at the boundary between order and chaos -- where fractal structures are most prominent. This "edge of chaos" hypothesis, explored by Chris Langton and Stuart Kauffman at the Santa Fe Institute, suggests that life itself may operate at this critical boundary.
- 21 / 23
Slide 22
Open Problems and Frontiers
- Despite decades of progress, fractal geometry remains an active research area with deep unsolved problems.
- MLC Conjecture
- Is the Mandelbrot set locally connected? This is the central open problem in holomorphic dynamics. A positive answer would imply that the Mandelbrot set is "topologically simple" despite its visual complexity. Partial results by Yoccoz (Fields Medal, 1994).
- Dimension of Brownian Motion Boundaries
- The frontier (outer boundary) of a planar Brownian motion has dimension 4/3 (proven by Lawler, Schramm, Werner in 2001, leading to a Fields Medal). But many related random fractal dimensions remain unproven.
- 3D Mandelbrot Analog
- There is no natural 3D analog of complex number multiplication. The Mandelbulb and Mandelbox are approximations using spherical coordinates, but no true 3D Mandelbrot set has been found. Quaternions and octonions produce interesting but different objects.
- Fractal Geometry of Quantum Mechanics
- Feynman path integrals suggest that quantum particle paths are fractal (D = 2). Can fractal geometry provide new insights into quantum gravity? The fractal structure of spacetime is an active area at the intersection of physics and mathematics.
- 22 / 23
Slide 23
Infinite Within the Finite
- Fractal geometry revealed that infinite complexity can emerge from the simplest of rules. It gave mathematics a language for the rough, the broken, the tangled -- for the shapes of the real world.
- 1.262Koch snowflake dimension
- 2.000Mandelbrot set boundary
- ∞Levels of detail
- "Bottomless wonders spring from simple rules, which are repeated without end."
- -- Benoit Mandelbrot
- 23 / 23
