Ensemble Learning with an Adversarial Hypergraph Model and a Convolutional Neural Network to Forecast Stock Price Variations
Bibliographic record
Abstract
AI has increased scholarly interest in predicting financial stock prices, a tough task.Recurrent neural network (RNN) time series price movement analysis is common but ignores market, investor, and headline factors.Graph neural networks excel at capturing complicated relationships and learning representations, making them useful for a variety of applications.This study investigates the predictive capabilities of graph neural networks about stock prices.Recent studies motivated to utilize hypergraphs for capturing intricate group-level data, such as corporate mobility patterns.Using a graph model to obtain paired linkages sets us apart from previous studies on hypergraphs.Also, demonstrated that using RNNs after applying this fundamental graph model is not advisable, in contrast to previous research.This study demonstrates the potential of future Recurrent Neural Networks (RNNs) to acquire knowledge about the long-term interconnections across companies, leading to enhanced predictive capabilities.The proposed model is an innovative ensemble learning framework that created specifically for the purpose of forecasting stock values.Part, one consists of a Graph Convolution Network (GCN), which encodes price and industry pairs.Part two involves a hypergraph convolution network that conducts adversarial training by modifying inputs before the final prediction layer.This modification facilitates the transmission of group-oriented information through hyperedges.There is empirical evidence that suggests that the proposed model outperforms approaches that are considered to be state-of-the-art on average and demonstrates remarkable performance during times when the market is declining.The proposed framework beats the second-best baseline by increments of 1.14%, 3.63%, and 2.45% for three different five-days, ten days and twenty days trade periods.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".