Paper: arXiv 2411.05790

Authors: Jue Xiao, Tingting Deng, Shuochen Bi

Abstract

In recent fast-paced financial markets, investors constantly seek ways to gain an edge and make informed decisions. Although achieving perfect accuracy in stock price predictions remains elusive, artificial intelligence (AI) advancements have significantly enhanced our ability to analyze historical data and identify potential trends. This paper takes AI driven stock price trend prediction as the core research, makes a model training data set of famous Tesla cars from 2015 to 2024, and compares LSTM, GRU, and Transformer Models. The analysis is more consistent with the model of stock trend prediction, and the experimental results show that the accuracy of the LSTM model is 94%. These methods ultimately allow investors to make more informed decisions and gain a clearer insight into market behaviors.

Complexity vs Empirical Score

  • Math Complexity: 5.5/10
  • Empirical Rigor: 6.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper uses standard deep learning architectures (LSTM, GRU, Transformer) with minimal novel derivations, placing it in the moderate math range. However, it includes a specific implementation on Tesla stock data (2015-2024) with reported accuracy metrics, indicating practical application and data-driven results.

Research Flowchart

  flowchart TD
  A["Research Goal: Compare AI Models for Stock Price Prediction"] --> B["Data: Tesla Stock 2015-2024"]
  B --> C["Model Training & Evaluation<br>LSTM vs. GRU vs. Transformer"]
  C --> D["Key Finding: LSTM achieved 94% Accuracy"]
  D --> E["Outcome: Informed Investment Decisions"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes to compare LSTM, GRU, and Transformer models for Tesla stock price prediction.

Source sentence
compares LSTM, GRU, and Transformer Models.

Assumes 3

  • The dataset has no missing values. data
    Source sentence
    the dataset covers 2,274 trading days with no missing values.
  • Differencing the series yields a stationary series suitable for modelling. statistical
    Source sentence
    Differencing the series improved stationarity (ADF test p-value: 0.142), enhancing its suitability for modelling.
  • LSTM is more suitable for stock time series than other models. model
    Source sentence
    From the perspective of overall indicators, the LSTM model is more suitable for stock time series

Data

Backtest sample: Tesla stock · equities

  • Tesla stock dataset · Tesla official website market-prices public
    2015-01-01 to 2024-01-16 · Tesla stock
    Source sentence
    The data set used in this study comes from Tesla's official website, which covers product information, sales data, and user reviews published in multiple years.

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • The LSTM model achieved 94% accuracy in predicting Tesla stock price trends. predictive-accuracy
    Source sentence
    the accuracy of the LSTM model is 94%.
  • The Transformer model achieved a fitting degree of 94.4%, which is 4.5% higher than the GRU model. performance
    Source sentence
    the fitting degree of the Transformer model reaches 94.4%, which is 4.5% higher than that of the GRU neural network.
  • The GRU model’s MAE, MSE and RMSE values are 14.73077754, 339.9002233 and 18.43638314 respectively, indicating reduced prediction error compared to baseline. statistical
    Source sentence
    Compared with the GRU model, the MAE value, MSE value and RMSE value are 14.73077754, 339.9002233 and 18.43638314, respectively, and the prediction error score of GRU model is significantly reduced.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 57 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.