Paper: arXiv 2404.07969

Authors: Chufeng Li, Jianyong Chen

Abstract

As a branch of time series forecasting, stock movement forecasting is one of the challenging problems for investors and researchers. Since Transformer was introduced to analyze financial data, many researchers have dedicated themselves to forecasting stock movement using Transformer or attention mechanisms. However, existing research mostly focuses on individual stock information but ignores stock market information and high noise in stock data. In this paper, we propose a novel method using the attention mechanism in which both stock market information and individual stock information are considered. Meanwhile, we propose a novel EMD-based algorithm for reducing short-term noise in stock data. Two randomly selected exchange-traded funds (ETFs) spanning over ten years from US stock markets are used to demonstrate the superior performance of the proposed attention-based method. The experimental analysis demonstrates that the proposed attention-based method significantly outperforms other state-of-the-art baselines. Code is available at https://github.com/DurandalLee/ACEFormer.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10 (capped from 7.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 7.0)

Why this score: The paper employs advanced deep learning architectures like Transformers and a novel signal processing algorithm (ACEEMD) with mathematical derivations, indicating high mathematical complexity. It also provides empirical validation using real financial data (two ETFs over ten years) and compares against state-of-the-art baselines, with code available, showing substantial empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Improve Stock Movement Forecasting"] --> B["Data Preparation<br>Two ETFs (10+ years)"]
  B --> C["Proposed Method<br>ACEFormer"]
  subgraph C_Method ["ACEFormer Components"]
      C1["Novel Position Mechanism"]
      C2["Improved EMD<br>Denoising"]
      C3["Transformer Attention<br>Market + Individual Info"]
  end
  C --> D["Computational Process<br>Model Training & Evaluation"]
  D --> E{"Performance Comparison"}
  E --> F["Outcome<br>Outperforms SOTA Baselines"]
  E --> G["Outcome<br>Superior ETF Forecasting"]

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 a stock trend prediction solution called ACEFormer, which includes a novel attention mechanism, a time-aware mechanism, and an improved EMD-based denoising algorithm (ACEEMD).

Source sentence
We propose a stock trend prediction solution called ACEFormer. It consists of

Assumes

No explicit assumptions found in the text.

Data

  • NASDAQ100 market-prices public
    spanning over ten years · 102 equity stocks of non‑financial companies from the NASDAQ
    Source sentence
    We evaluate the proposed method on two real‑world datasets, which are NASDAQ100 and SPY500 [24], from US stock markets spanning over ten years.
  • SPY500 market-prices public
    spanning over ten years · Standard and Poor’s 500, tracking the stock performance of 500 large companies listed on U
    Source sentence
    We evaluate the proposed method on two real‑world datasets, which are NASDAQ100 and SPY500 [24], from US stock markets spanning over ten years.

Costs

Not modeled A trading or execution result is reported without costs.

Finds 2

  • ACEFormer achieves the highest trend prediction performance, with ACC of 69.23% and MCC of 0.379 on SPY500 (and 69.23% and 0.382 on NASDAQ100), outperforming all baselines. performance
    Source sentence
    seen that among all experimental models, ACEFormer performs the best. In terms of trend evaluation metrics, the ACC and MCC results of ACEFormer on the SPY500 (NASDAQ100) dataset are 69.23% and 0.379 (69.23% and 0.382), respectively.
  • ACEEMD effectively reduces short‑term noise while preserving critical turning points and avoids endpoint effects. predictive-accuracy
    Source sentence
    ACEEMD not only solves the endpoint effect problem but also avoids over‑denoising and effectively keeps key turning points in stock trends.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 13 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

https://github.com/DurandalLee/ACEFormer — the authors' own repository

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

Cited References (23)

  1. Exploring the impact of investor’s sentiment tendency in varying input window length for stock price prediction (2023) — link
  2. TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis (2022) — link
  3. Non-stationary Transformers: Exploring the Stationarity in Time Series Forecasting (2022) — link
  4. Coal Thickness Prediction Method Based on VMD and LSTM (2022) — link
  5. Stock Selection via Spatiotemporal Hypergraph Attention Network: A Learning to Rank Approach (2021) — link
  6. Prediction of Trends in Stock Market using Moving Averages and Machine Learning (2021) — link
  7. A Deep Neural Network Based Model For Stock Market Prediction (2021) — link
  8. Prediction of Stock Closing Prices Based on Attention Mechanism (2020) — link
and 15 more
  1. Enhancing Stock Movement Prediction with Adversarial Training (2018) — link
  2. An Adaptive SVR for High-Frequency Stock Price Forecasting (2018) — link
  3. Attention is All you Need (2017) — link
  4. Novel and efficient computation of Hilbert-Huang transform on surfaces (2016) — link
  5. Neural Machine Translation by Jointly Learning to Align and Translate (2014) — link
  6. Empirical Mode Decomposition-Based Time-Frequency Analysis of Multivariate Signals: The Power of Adaptive Data Analysis (2013) — link
  7. A Chaotic Attractor For the S&P 500 (1991) — link
  8. MSGraph: Modeling multi-scale K-line sequences with graph attention network for profitable indices recommendation (2023) — link
  9. A Dual-Staged Attention Based Conversion-Gated Long Short Term Memory for Multivariable Time Series Prediction (2022) — link
  10. An Effective Source Number Enumeration Approach Based on SEMD (2022) — link
  11. Informer: Beyond efficient transformer for long sequence time-series forecasting (2021)
  12. Predicting Stock Market Trends Using Machine Learning and Deep Learning Algorithms Via Continuous and Binary Data; a Comparative Analysis (2020) — link
  13. Hierarchical multi-scale gaussian transformer for stock movement prediction (2020)
  14. Computational intelligence and financial markets: a survey and future directions (2016)
  15. Comparative study of stock trend prediction using time delay, recurrent and probabilistic neural networks (1998)

0 of 23 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.