Hybrid Deep Learning Architectures for Stock Market Prediction
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Accurate stock market prediction is a challenging task due to the volatile and nonlinear nature of it which depends on numerous factors as local and global economic conditions, company specific performance etc.It is not possible to account all existing relevant factors which influence the stock market in order to make respective trading decisions without having appropriate algorithms and techniques.A recent development of deep learning for making trading decisions has been growing rapidly with numerous research papers addressing stock market forecasting as time series regression problem.In recent years, Long Short-Term Memory (LSTM) neural networks have become the state-of-the-art models for a variety of machine learning problems which differ significantly in scale and nature.The central idea behind the LSTM architecture is a memory cell which can maintain its state over time, and non-linear gating units which regulate the information flow into and out of the cell.Most modern studies incorporate many improvements that have been made to the LSTM architecture since its original formulation.Considering the complexity of financial time series, combining deep learning with financial market prediction is regarded as a very important topic of research.The experiments in this study are divided into two sets which use different topologies.For our first experimental set we created the following hybrid sequential nonlinear models Convolutional Neural Networks (CNN), stacked LSTM, BiDirectional LSTM (BiLSTM) and Gated Recurrent Unit (GRU) with BiLSTM.For our second experimental set we propose a new deep learning topology based on Attention CNN_BiLSTM for pretraining and Light Gradient Bosting Machine (LGBM) as a regressor.The evaluation of experimental results indicates that the last proposed model achieves better performance in predicting stock market when compared to the models proposed in the first experimental set.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it