Machine learning program

AQUILES

Data Mining Rule Induction & Extraction Reloaded

VTAlgo's 2026 flagship: induce robust, explainable trading rules from data using the newest AI techniques — with the same anti-over-optimization discipline that governs AEGIS.

$1,800

Selective admission 5 phases · 5 modules + Validation layer Live cohort English
Curriculum · 5 phases · 5 modules + Validation layer
Phase 1
Data Engineering
PIT · gaps · quality
Phase 2
Sampling / Bars
Time · volume · dollar
Phase 3
Feature Engineering
Price · flow · regime · path
Phase 4
Feature Selection
IC · FDR · stability
Phase 5
Rule Discovery
Induction · RuleFit · ensembles
Validation · transversal layer · not a sixth module
Accompanies Data → Sampling → Features → Selection → Rules. Standing controls: leakage → temporal integrity → multiple testing → stability → OOS.
Session counts and hours are not sealed. Lesson bodies are not authored.

What you'll learn

  • Engineer intradaily data under PIT discipline (timestamps, gaps, multi-asset sync, quality)
  • Build comparable bars — time, volume, and dollar — and compare sampling schemes
  • Engineer features across price, flow, structure, regime, cross-asset, and path
  • Select features under redundancy, dependence, IC, multicollinearity, FDR, and temporal stability
  • Discover explainable rules (induction, extraction, association, subgroup, sequential, RuleFit, ensembles)
  • Carry validation through every module: leakage → temporal integrity → multiple testing → stability → OOS

Program content

Phase 1 — Data Engineering 1 module · 6 chapters
Module 1 — Data Engineering Validation layer runs here
Intraday data
PIT / timestamps
Cleaning
Gaps
Multi-asset synchronization
Data quality
Phase 2 — Market Sampling / Bar Construction 1 module · 4 chapters
Module 2 — Market Sampling / Bar Construction Validation layer runs here
Time bars: 1m, 5m, 15m, 30m, 60m
Volume bars
Dollar bars
Sampling-scheme comparison
Phase 3 — Feature Engineering 1 module · 9 chapters
Module 3 — Feature Engineering Validation layer runs here
Price / returns
Volatility
Volume / flow
Market structure
Temporal
Cross-asset
Cross-sectional
Regime
Path / sequence
Phase 4 — Feature Selection 1 module · 8 chapters
Module 4 — Feature Selection Validation layer runs here
Redundancy
Dependence
IC / predictive information
Stability
Multicollinearity
Univariate / multivariate selection
Multiple testing / FDR
Temporal stability
Phase 5 — Rule Discovery 1 module · 9 chapters
Module 5 — Rule Discovery Validation layer runs here
Rule Induction
Rule Extraction
Association Rules
Subgroup Discovery
Sequential Pattern Mining
Emerging Patterns
Trees → rules
RuleFit
Rule Ensemble

Course content is Phase → Module → Chapter. Five phases, each with one module. Validation is a transversal research layer through Data → Sampling → Features → Selection → Rules, not a sixth phase. Lesson bodies are not authored. Session counts and hours are not sealed.

Description

AQUILES (formerly DMRI-2026) is VTAlgo's flagship 2026 program on Data Mining: Rule Induction & Extraction. It teaches how to induce robust, explainable trading rules from data with the latest AI, while refusing the shortcut of over-optimization. Validation is not a final chapter: it accompanies every module.

AQUILES is a paid program. It is not included with the ZEUS Engine.

Instructor

Elmer Niño

Elmer Niño

Founder, VTAlgo Group · Systems Engineer, MSc Computer Science

Algorithmic trader since 2009, running trading as an enterprise. Methodology built on walk-forward analysis, genetic optimization, PCA against over-optimization, the Triple Barrier Method, and 9–12 month live validation before capital.