volarixs - applied AI & ML to finance

Explore our latest posts on machine learning, market dynamics, strategy architecture and design

Feature Engineering

Shrinking the Feature Space: PCA & Autoencoders

Many features are redundant or noisy. High dimensionality = harder to generalize.

How Asset Managers Can Implement AI & Machine Learning

Part 2: Infrastructure, Governance & Roadmap. What it takes to implement AI in asset management.

AI Implementation

Neural Networks for Market Data: MLPs, CNNs & LSTMs

We are selective with deep learning. Expensive to train, easy to overfit, harder to debug.

Neural Networks

Signal Half-Life and Decay: How Long Do ML Edges Really Last?

If you discover a signal today, how long will it work?

Edge Persistence

How Asset Managers Can Use AI & Machine Learning in Investment Decisions

Part 1: Use Cases & Value. Real-world use cases: idea generation, regime analysis, risk management.

Asset Management

Modeling Market Turbulence: GARCH, EGARCH & HAR

Volatility ≠ returns: heavy tails, clustering, mean reversion. Dedicated volatility models are essential.

ARIMA, SARIMAX & VAR: When Classical Time-Series Still Win

Explicitly model temporal dependence with transparent structure.

Volatility Forecasting Benchmarks: GARCH, HAR, and ML

Compare GARCH, HAR, and ML models for volatility forecasting.

Machine Learning

How Market Regimes Break ML Models

Financial machine learning rarely fails because the model is 'bad'. It fails because the market regime changed.

Boosted Trees for Alpha: XGBoost & LightGBM

Gradient boosting dominates tabular ML. Learn how XGBoost and LightGBM deliver strong performance.

The 19 Most Important Features for Equity Return Forecasting

Most ML performance in finance doesn't come from the model — it comes from the features.

Rolling Windows for Financial ML: A Complete Guide

If you use financial data and your model does not use a rolling window, the backtest is wrong.

Rolling Windows

Beyond Sharpe: A Research Framework for Evaluating ML Trading Strategies

Sharpe ratio is dangerously incomplete for ML strategies.

Random Forests in Finance: Nonlinear Signals Without the Drama

Tree-based ensembles capture nonlinearities and interactions in market data.

From Linear Regression to Lasso: Fast, Interpretable Baselines

Linear and regularized regressions still do serious work in finance.

Linear Regression

Market Regimes, Clusters & HMMs: Teaching Models to Respect the Environment

Episodes where statistical properties are stable enough: high vol vs low vol, risk-on vs risk-off.

Building a Universe-Wide Prediction Grid

An alpha factory needs predictions for every asset at multiple horizons from multiple models.

Prediction Grid

Regime-Conditioned Performance: Measuring ML Robustness

Most backtests report a single Sharpe. But ML models fail by regime.

volarixs - applied AI & ML to finance

Explore our latest posts on machine learning, market dynamics, strategy architecture and design

Feature Engineering

Shrinking the Feature Space: PCA & Autoencoders

Many features are redundant or noisy. High dimensionality = harder to generalize.

How Asset Managers Can Implement AI & Machine Learning

Part 2: Infrastructure, Governance & Roadmap. What it takes to implement AI in asset management.

AI Implementation

Neural Networks for Market Data: MLPs, CNNs & LSTMs

We are selective with deep learning. Expensive to train, easy to overfit, harder to debug.

Neural Networks

Signal Half-Life and Decay: How Long Do ML Edges Really Last?

If you discover a signal today, how long will it work?

Edge Persistence

How Asset Managers Can Use AI & Machine Learning in Investment Decisions

Part 1: Use Cases & Value. Real-world use cases: idea generation, regime analysis, risk management.

Asset Management

Modeling Market Turbulence: GARCH, EGARCH & HAR

Volatility ≠ returns: heavy tails, clustering, mean reversion. Dedicated volatility models are essential.

ARIMA, SARIMAX & VAR: When Classical Time-Series Still Win

Explicitly model temporal dependence with transparent structure.

Volatility Forecasting Benchmarks: GARCH, HAR, and ML

Compare GARCH, HAR, and ML models for volatility forecasting.

Machine Learning

How Market Regimes Break ML Models

Financial machine learning rarely fails because the model is 'bad'. It fails because the market regime changed.

Boosted Trees for Alpha: XGBoost & LightGBM

Gradient boosting dominates tabular ML. Learn how XGBoost and LightGBM deliver strong performance.

The 19 Most Important Features for Equity Return Forecasting

Most ML performance in finance doesn't come from the model — it comes from the features.

Rolling Windows for Financial ML: A Complete Guide

If you use financial data and your model does not use a rolling window, the backtest is wrong.

Rolling Windows

Beyond Sharpe: A Research Framework for Evaluating ML Trading Strategies

Sharpe ratio is dangerously incomplete for ML strategies.

Random Forests in Finance: Nonlinear Signals Without the Drama

Tree-based ensembles capture nonlinearities and interactions in market data.

From Linear Regression to Lasso: Fast, Interpretable Baselines

Linear and regularized regressions still do serious work in finance.

Linear Regression

Market Regimes, Clusters & HMMs: Teaching Models to Respect the Environment

Episodes where statistical properties are stable enough: high vol vs low vol, risk-on vs risk-off.

Building a Universe-Wide Prediction Grid

An alpha factory needs predictions for every asset at multiple horizons from multiple models.

Prediction Grid

Regime-Conditioned Performance: Measuring ML Robustness

Most backtests report a single Sharpe. But ML models fail by regime.

Feature Engineering

June 2, 2026

9 min read

Shrinking the Feature Space: PCA & Autoencoders for Market Data

Many features are redundant or noisy. High dimensionality = harder to generalize. We often want a compact factor representation.

1. Why Dimensionality Reduction

Many features are redundant or noisy.
High dimensionality = harder to generalize.
We often want a compact factor representation.

2. PCA (Principal Component Analysis)

Finds directions of maximum variance. In finance, first few components often look like:

Market factor
Sector or style tilts

Great for yield curves and cross-sectional equity returns.

3. Autoencoders (Roadmap Feature)

Neural networks that learn compressed representation. Nonlinear analogue of PCA.

Potentially capture structures PCA misses.

4. How This Fits volarixs

The platform is organised around feature sets that feed the model classes you pick in an experiment — regression, trees and boosted models, neural networks, and time-series. Dimensionality reduction is the lens we apply to those feature sets, with PCA as the workhorse:

Compress correlated features into a few components before they reach a model.
Hand those components to a downstream model (linear, trees, or boosters) in place of the raw inputs.

Autoencoders sit further out on the roadmap, aimed at complex multi-modal feature sets (prices + fundamentals + text scores) where a nonlinear encoder can earn its keep over PCA.

PCA

Autoencoders

Dimensionality Reduction

Get new research in your inbox

Applied AI & ML for the buy-side — new research on signals, regimes, and strategy design, straight to your inbox. No noise.

Ready to compress your features?

Build experiments on compact feature sets in volarixs.

Explore More Research

Feature Engineering

June 2, 2026

9 min read

Shrinking the Feature Space: PCA & Autoencoders for Market Data

Many features are redundant or noisy. High dimensionality = harder to generalize. We often want a compact factor representation.

1. Why Dimensionality Reduction

Many features are redundant or noisy.
High dimensionality = harder to generalize.
We often want a compact factor representation.

2. PCA (Principal Component Analysis)

Finds directions of maximum variance. In finance, first few components often look like:

Market factor
Sector or style tilts

Great for yield curves and cross-sectional equity returns.

3. Autoencoders (Roadmap Feature)

Neural networks that learn compressed representation. Nonlinear analogue of PCA.

Potentially capture structures PCA misses.

4. How This Fits volarixs

The platform is organised around feature sets that feed the model classes you pick in an experiment — regression, trees and boosted models, neural networks, and time-series. Dimensionality reduction is the lens we apply to those feature sets, with PCA as the workhorse:

Compress correlated features into a few components before they reach a model.
Hand those components to a downstream model (linear, trees, or boosters) in place of the raw inputs.

Autoencoders sit further out on the roadmap, aimed at complex multi-modal feature sets (prices + fundamentals + text scores) where a nonlinear encoder can earn its keep over PCA.

PCA

Autoencoders

Dimensionality Reduction

Get new research in your inbox

Applied AI & ML for the buy-side — new research on signals, regimes, and strategy design, straight to your inbox. No noise.

Ready to compress your features?

Build experiments on compact feature sets in volarixs.

Explore More Research