quantum-probability-statistics - SKILL.md Agent Skill

name: quantum-probability-statistics description: "Framework for applying quantum probability theory to statistical settings and machine learning. Covers Born rule applications, quantum measurement theory, quantum state superposition, and quantum interference in probabilistic modeling. Activation: quantum probability, quantum statistics, 量子概率统计, quantum ML, Born rule statistics."

Quantum Probability Statistics

Framework for applying quantum probability theory to classical statistical settings, particularly machine learning and probabilistic modeling.

Description

Quantum probability offers novel approaches to statistical problems through:

Born rule: Probability as squared amplitude (new interpretation of uncertainty)
Quantum superposition: Multi-modal probability distributions
Quantum entanglement: Correlation modeling beyond classical limits
Quantum interference: Probability fusion with constructive/destructive effects

Use when:

Modeling complex multi-modal distributions
Capturing correlations that exceed classical bounds
Applying Born rule to statistical inference
Quantum-inspired machine learning architectures

Activation Keywords

quantum probability
quantum statistics
量子概率统计
quantum machine learning
quantum probability applications
Born rule statistics
quantum Bayesian inference
quantum Monte Carlo

Tools Used

exec: Run quantum probability calculations, simulations
read: Load quantum theory references, paper summaries
write: Generate quantum probability analysis reports
web_fetch: Fetch latest quantum probability research from arxiv

Core Concepts

Born Rule in Statistics

The Born rule states that the probability of measuring a quantum state is the squared norm of its amplitude:

P(state) = |ψ|²

Statistical interpretation:

Amplitude represents "potential" or "pre-probability"
Squaring creates actual probability
Enables interference effects (constructive/destructive)

Quantum State as Probability Distribution

Quantum state |ψ⟩ = α₁|₁⟩ + α₂|₂⟩ + ... + αₙ|n⟩

Statistical analogy:

Superposition = Multi-modal distribution
Amplitudes αᵢ = Distribution weights (before normalization)
Measurement = Sampling from distribution
Collapse = Realization of one outcome

Quantum Entanglement for Correlations

Entangled state: |ψ⟩ = α|AB⟩ + β|A'B'⟩

Correlation modeling:

Stronger correlations than classical probability allows
Bell inequality violations → beyond classical limits
Applications: Correlation matrices with quantum-enhanced bounds

Quantum Interference

Probability amplitude interference:

ψ_total = ψ₁ + ψ₂
P_total = |ψ₁ + ψ₂|² = |ψ₁|² + |ψ₂|² + 2·Re(ψ₁*·ψ₂)

Statistical application:

Interference term: constructive (enhance) or destructive (suppress)
Probability fusion beyond simple averaging
Context-dependent probability adjustment

Usage Patterns

Pattern 1: Quantum Bayesian Inference

Replace classical Bayesian update with quantum probability update:

# Classical: P(H|D) = P(D|H)·P(H) / P(D)
# Quantum: |ψ_H⟩' = P(D|H)·|ψ_H⟩ / √P(D)

# Quantum advantage: preserves phase information
# Enables interference in sequential updates

Pattern 2: Multi-modal Distribution Modeling

Use quantum superposition for multi-modal distributions:

# Classical: P(x) = w₁·P₁(x) + w₂·P₂(x)
# Quantum: |ψ(x)⟩ = √w₁·|ψ₁(x)⟩ + √w₂·|ψ₂(x)⟩

# Quantum advantage: interference between modes
# Better representation of uncertainty

Pattern 3: Quantum Monte Carlo

Quantum-enhanced sampling:

# Use quantum circuits to generate samples
# Quantum random walks for exploration
# Quantum tunneling for escaping local minima

Instructions for Agents

Step 1: Identify Application Context

Determine if quantum probability offers advantages:

Multi-modal uncertainty?
Strong correlations beyond classical bounds?
Context-dependent probability adjustments?
Sequential updates needing interference effects?

Step 2: Map to Quantum Formalism

Translate statistical problem to quantum language:

Random variables → Quantum observables
Probability distribution → Quantum state
Correlation → Entanglement
Bayesian update → Quantum measurement

Step 3: Apply Quantum Operations

Execute quantum-inspired operations:

Superposition for multi-modal modeling
Entanglement for correlation modeling
Interference for probability fusion
Measurement for outcome realization

Step 4: Convert Back to Statistics

Map quantum results back to statistical interpretation:

Compute probabilities via Born rule
Interpret interference effects
Analyze quantum-enhanced correlations

Step 5: Validate and Report

Validate quantum probability results:

Compare with classical baselines
Check physical consistency (Born rule)
Document quantum advantages

Mathematical Foundation

Key Equations

Born Rule:

P(measurement = i) = |⟨i|ψ⟩|²

Quantum State Evolution:

|ψ(t)⟩ = U(t)|ψ(0)⟩  (unitary evolution)

Quantum Entanglement Metric:

Concurrence C = max(0, λ₁ - λ₂ - λ₃ - λ₄)
where λᵢ are eigenvalues of √(ρ·ρ̃·ρ)

Quantum Interference:

I = 2·Re(ψ₁*·ψ₂)  (interference term)

Applications in Machine Learning

1. Quantum-inspired Neural Networks

Quantum probability layers
Amplitude-based activation functions
Entanglement for feature correlations

2. Quantum Bayesian Networks

Quantum DAG structures
Born rule for conditional probabilities
Quantum message passing

3. Quantum Reinforcement Learning

Quantum state as belief state
Quantum exploration strategies
Quantum value functions

4. Quantum Anomaly Detection

Quantum measurement deviations
Born rule anomaly scoring
Quantum interference anomalies

References

Core Papers

Quantum probability for statisticians; some new ideas (arxiv:2503.02658)
- Born rule arguments
- Statistical applications of quantum probability
- Machine learning applications list
Quantum probability as a theory of decision making
- Quantum cognition models
- Behavioral economics applications
Quantum Machine Learning (Schuld, Sinayskiy, Petruccione)
- Quantum ML foundations
- Quantum probability in ML

External Resources

Error Handling

Phase Inconsistency

If quantum phase information lost:

Reconstruct phase from correlation data
Use maximum entropy principle
Apply phase retrieval algorithms

Non-physical Probabilities

If Born rule yields invalid probabilities:

Check state normalization
Verify unitary operations
Correct computational errors

Classical Limit Violation

If quantum correlations exceed physical limits:

Validate Bell inequality bounds
Check measurement consistency
Apply quantum decoherence corrections

Examples

Example 1: Quantum Bayesian Update

# Classical Bayes: P(H|D) = P(D|H)·P(H) / P(D)

# Quantum Bayes: preserves amplitude and phase
def quantum_bayes_update(prior_state, likelihood, evidence):
    """
    Quantum Bayesian inference preserving phase information.
    
    prior_state: |ψ_H⟩ (amplitude + phase)
    likelihood: P(D|H) (measurement operator)
    evidence: √P(D) (normalization)
    
    Returns: Updated quantum state |ψ_H⟩'
    """
    updated = likelihood * prior_state / evidence
    return updated  # Contains phase for interference

Example 2: Multi-modal with Quantum Superposition

# Classical mixture: P(x) = w₁·P₁(x) + w₂·P₂(x)

# Quantum superposition: enables interference
def quantum_multimodal(modes, weights, x):
    """
    Quantum superposition for multi-modal distribution.
    
    modes: List of quantum states |ψ₁⟩, |ψ₂⟩, ...
    weights: Amplitudes √w₁, √w₂, ...
    x: Measurement point
    
    Returns: Probability with interference effects
    """
    psi_total = sum(w * m for w, m in zip(weights, modes))
    probability = abs(psi_total)**2
    return probability  # Born rule with interference

Example 3: Quantum Correlation Matrix

# Classical correlation: bounded by [-1, 1]
# Quantum correlation: can exceed classical bounds

def quantum_correlation_matrix(features):
    """
    Quantum-enhanced correlation modeling.
    
    Uses entangled states for feature correlations.
    Enables correlations beyond classical limits.
    """
    # Create entangled state for feature pairs
    # Compute quantum correlation via Bell inequality
    # Return enhanced correlation matrix
    pass

Related Skills

quantum-statistical-estimation: Quantum parameter estimation
quantum-tensor-network-ml: Tensor networks for quantum ML
quantum-geometric-statistical-analysis: Quantum Fisher information geometry

Notes

Quantum probability ≠ quantum computing (no quantum hardware needed)
Mathematical framework only (can simulate on classical computers)
Key advantage: phase information enables interference effects
Born rule is central: probability = |amplitude|²
Applications growing in ML, decision theory, cognitive science

Created from paper: Quantum probability for statisticians; some new ideas (arxiv:2503.02658) Date: 2026-04-10 Source: Weekly research task (Friday: Mathematics + Quantum)