Optimal Text-Based Time-Series Indices

By David Ardia, Keven Bluteau

Rating

1413
Battle Count: 70

Relevance

8/10
Highly relevant for creating custom indicators from news data for trading strategies

Implementation Complexity

7/10
Requires implementation of custom genetic algorithm and access to large news corpus

Reproducibility

4/5
Detailed methodology provided, but full code not available

About this paper

Methodology: Genetic Algorithm for Text Selection. Problem types: Time Series Forecasting, Natural Language Processing, Optimization.

The interactive Everscope explorer (charts, battles, favorites) loads below.