Risk-Kit 🎯

What: Expert-driven credit risk scorecards with native sklearn compatibility

Why: Combines domain expertise with ML best practices - interpretable, auditable, and production-ready

How: Define scoring rules, get sklearn estimator

from risk_kit import ExpertScorecard, NumericFeature, NumericBucket
from risk_kit.expert_scorecard.validation import ValidatorRegistry, FeatureWeightValidator

# Create validation registry
registry = ValidatorRegistry()
registry.register(FeatureWeightValidator)

# Define expert scoring rules with validation
scorecard = ExpertScorecard(
    features=[
        NumericFeature(
            name="income",
            buckets=[
                NumericBucket(definition=(0, 50000), score=20),
                NumericBucket(definition=(50000, float('inf')), score=80)
            ],
            weight=100
        )
    ],
    validation_registry=registry
)

# Use like any sklearn model
scorecard.fit(X_train, y_train)
predictions = scorecard.predict(X_test)

# Pickle for production
import pickle
pickle.dump(scorecard, open('model.pkl', 'wb'))

Installation

pip install risk-kit

Core Concepts

Buckets → Features → Scorecard → Sklearn Model

Each feature has buckets that assign scores to value ranges. Features are weighted and combined into a scorecard that acts as a standard sklearn estimator.

Getting Started

1. Install: pip install risk-kit

2. Learn: Quick Start (5 minutes)

3. Example: Iris Dataset Example (complete workflow)

4. Visualize: Scorecard Visualization (scorecard tables)

5. Reference: API Reference (full API)