Enhancing Stock Price Forecasting with LSTM Networks: A Comparative Study of Classic and PSO-Optimized Models on S&P 500 Stocks
Bibliographic record
Abstract
This study is part of an empirical and quantitative approach aimed at improving stock market fluctuation forecasting through the application of artificial intelligence models. More specifically, it evaluates the performance of two methods based on Long Short-Term Memory (LSTM) neural networks, one of the most powerful algorithms for analyzing financial time series. The first method is grounded in a classic LSTM model, while the second incorporates hyperparameter optimization using the Particle Swarm Optimization (PSO) metaheuristic method, allowing for better convergence and enhanced prediction accuracy. The study is conducted on ten stocks representing the US S&P 500 index, with historical data spanning several decades, collected via the Investing.com and Yahoo Finance platforms. The empirical results demonstrate a clear superiority of the LSTM-PSO model regarding predictive accuracy, with significant reductions in errors (MSE, RMSE, MAE, MSLE, and RMSLE) compared to the traditional model. These findings emphasize the advantages of combining artificial intelligence and algorithmic optimization for handling complex financial data. In the global context of digitization and automation of investment decisions, this research contributes significantly to the development of reliable predictive systems. Finally, the study raises the question of whether this methodological framework could be effectively adapted to emerging markets, such as the Moroccan Stock Market, where financial environments are characterized by lower trading volumes, different volatility patterns, and more limited historical data. This opens up avenues for future research into the challenges and opportunities of applying advanced AI-based forecasting models in less mature financial markets.
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 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 | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".