failed_experiment.db
← Back to browse

Decision Sciences› Management Science and Operations Research› Stock Market Forecasting Methods

Financial forecasting — Neural networks and SVMs forecasting KOSPI 200 trends

Submitted 2026-10-22

Field: Decision Sciences › Management Science and Operations ResearchType: No effect foundArea: financial forecastingTarget: Neural networks and SVMs forecasting KOSPI 200 trendsStage: full study

Hypothesis

Nonparametric machine-learning models — neural networks and support vector machines with polynomial and RBF kernels — would forecast the direction of Korea's KOSPI 200 index, with Google Trends search data and ensembling improving accuracy.

What failed

The results were inconsistent with prior studies generally credited with high prediction performance: Google Trends terms were not effective predictors in these frameworks, and ensemble methods did not improve accuracy.

Broken assumption

Machine-learning models, search-trend features and ensembling can predict stock index direction as earlier studies claimed.

Why

The paper reports the hypothesis tests on these controversial points rather than a mechanism; the implication is that earlier reported performance does not hold up under its framework.

From the published paper: Pyo et al. PLoS ONE, 2017. ↗CC BY

Research ideas from this failure

Unvetted lead — read the papers

New hypotheses this failure suggests, each with a crucial experiment, a rival explanation of the original result, and the papers it rests on. Written by AI from paper titles and abstracts found on OpenAlex (and Europe PMC, for life and health sciences).

Related failures

Nothing else logged looks close to this one yet.

Links in published research

Where this experiment's genes are known to matter outside its own research area, and the papers behind those links. Sources: Open Targets Platform and Europe PMC.