skills.rest
ebrahim-sani
ebrahim-saniCOMMUNITY
Explore all skills by this author →
Website𝕏@iAmImranXGitHub

ml-strategy

Train machine-learning models on OHLCV data to generate trading signals.

Updated Apr 10, 2026
Search Tags:#signals#trading#ml#ohlcv#machine-learning#walk-forward
One-click install
npx skills add https://github.com/ebrahim-sani/trading-automation --skill ml-strategy-ebrahim-sani
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ml-strategy
Source: https://github.com/ebrahim-sani/trading-automation/tree/main/vibe-trading/agent/src/skills/ml-strategy
Command: npx skills add https://github.com/ebrahim-sani/trading-automation --skill ml-strategy-ebrahim-sani

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill builds predictive trading signals from OHLCV data using machine-learning models with walk-forward training to prevent data leakage and improve robustness.

Core Features & Use Cases

  • Walk-forward training: expands training data over time and retrains models to generate up-to-date signals.
  • Feature engineering: derives momentum, volatility, RSI-like indicators, moving-average ratios, and volume-derived features from OHLCV data.
  • Model versatility: supports RandomForest, GradientBoosting, and Ridge (logistic regression) classifiers for flexible risk/complexity trade-offs.
  • Use Case: traders can generate continuous signals in [-1.0, 1.0] indicating bearish to bullish expectations for each asset.

Quick Start

Provide your OHLCV dataset to the system and run the ml-strategy skill to generate signals for your assets.

Frequently Asked Questions about ml-strategy

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate trading signals from OHLCV data using machine learning?▼

To generate trading signals from OHLCV data, this skill trains machine-learning models like RandomForest and GradientBoosting on your dataset to output continuous signals in the [-1.0, 1.0] range for each asset.

What is walk-forward training in machine learning trading models?▼

Walk-forward training is a technique that expands training data over time and retrains machine-learning models sequentially, which prevents data leakage and improves robustness when generating predictive trading signals.

Can I use GradientBoosting or Ridge models on historical OHLCV data for signal generation?▼

Yes, you can apply GradientBoosting, Ridge, or RandomForest models to historical OHLCV data. The system supports these classifiers to provide flexible risk and complexity trade-offs for signal generation.

What features do I need in my OHLCV dataset for machine learning signal processing?▼

You need a clean OHLCV dataset containing open, high, low, close, and volume columns. The process derives momentum, volatility, RSI-like indicators, moving-average ratios, and volume features automatically.

Do I need scikit-learn and pandas to run machine learning trading signals?▼

Yes, you need Python libraries including scikit-learn, pandas, and numpy installed in your environment to process the OHLCV dataset, engineer features, and train the machine-learning models.

How does walk-forward training prevent data leakage in ML trading strategies?▼

Walk-forward training prevents data leakage by strictly expanding the training window over time and retraining models sequentially, ensuring that future data is never used to predict past trading signals.

Related Skills

Explore More Data & Analytics Skills→
kerrybackkerryback

momentum-strategy

Compute momentum signals, form quintile portfolios, and backtest long-short strategies.

Community
Intermediate
loanntcloanntc

news-event-strategy

Convert sentiment-scored news events into time-decayed CSV signals for backtesting.

Community
Intermediate
kmshihab7878kmshihab7878
6

finance-ml

Train LSTM, XGBoost, and transformer models for financial price prediction and backtesting.

Community
Advanced
20YN0420YN04

ml-strategy

Generate predictive trading signals from OHLCV data using machine learning models.

Community
Advanced