https://github.com/konvsys/quantitative-kinematics-trading

Institutional-grade market regime decoding via Savitzky-Golay Kinematics and Hidden Markov Models (HMM). Engineered for zero-lag signal demodulation and structural alpha detection.
https://github.com/konvsys/quantitative-kinematics-trading

hmm market-regime pinescript quantitative-finance signal-processing

Last synced: 4 months ago
JSON representation

Institutional-grade market regime decoding via Savitzky-Golay Kinematics and Hidden Markov Models (HMM). Engineered for zero-lag signal demodulation and structural alpha detection.

• Host: GitHub
• URL: https://github.com/konvsys/quantitative-kinematics-trading
• Owner: konvsys
• Created: 2026-02-28T15:06:43.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2026-02-28T17:12:50.000Z (4 months ago)
• Last Synced: 2026-02-28T18:33:37.147Z (4 months ago)
• Topics: hmm, market-regime, pinescript, quantitative-finance, signal-processing
• Language: Jupyter Notebook
• Homepage: https://konvsys.github.io/impulse-docs/
• Size: 625 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Quantitative Kinematics & Regime Decoding

![Impulse-X Terminal Preview](assets/Impulse_x_1.png)

## The Problem: The Phase Lag Death Spiral

Most systematic strategies fail because they rely on **time-averaging indicators** (Moving Averages, RSI, MACD). In a non-stationary system like the limit order book, averaging past data creates a terminal phase lag. By the time your indicator signals a "trend," the institutional move is already exhausted.

## The Solution: Market Physics

This repository demonstrates a **Kinematic approach** to price action. Instead of treating price as a series of static candles, we treat it as a continuous physical object in motion.

### Core Mathematical Pillars:

* **Savitzky-Golay Filtering:** Replacing lagging MAs with polynomial-fit derivatives to extract instantaneous **Velocity ($v$)** and **Acceleration ($a$)**.

* **Causal Hilbert Transforms:** Utilizing reflective padding to extract the instantaneous phase of the market cycle without "end-point repainting."

* **Shannon Entropy ($H$):** Measuring the Information Density of the signal to identify High Chaos environments where trend-following is statistically suicidal.

* **Hidden Markov Models (HMM):** A Bayesian framework that synthesizes Kinematics and Entropy to decode the hidden structural regime of the market (e.g., *Bull Drift* vs. *Kinematic Dip*).

---

## 🔬 High-Performance Research

The logic in this repo is a subset of the **Impulse-X Engine**, a high-speed terminal environment optimized for disciplined traders who require objective math over gut feelings.

### Performance Characteristics:

* **Strictly Causal:** No look-ahead bias. No repainting.

* **Signal Demodulation:** Separating structural alpha from high-entropy noise.

* **Regime Awareness:** Automatically adjusting bias based on the HMM transition matrix.

---

## 📘 Deep Dive Documentation

For the full mathematical derivation, LaTeX frameworks, and terminal implementation specs, visit the official hub:

👉 **[Impulse-X Documentation & Technical Manual](https://konvsys.github.io/impulse-docs/)**

---

📊 TradingView User? I've included a free Pine Script version of the Kinematic Velocity engine. Copy the code from the /pinescript folder and paste it into your editor.

[Secure your Alpha Seat on Gumroad](https://konvolut.gumroad.com/l/wjtou)