Paper: arXiv 2409.00480

Authors: Mohit Apte, Yashodhara Haribhakta

Abstract

In the rapidly evolving field of financial forecasting, the application of neural networks presents a compelling advancement over traditional statistical models. This research paper explores the effectiveness of two specific neural forecasting models, N-HiTS and N-BEATS, in predicting financial market trends. Through a systematic comparison with conventional models, this study demonstrates the superior predictive capabilities of neural approaches, particularly in handling the non-linear dynamics and complex patterns inherent in financial time series data. The results indicate that N-HiTS and N-BEATS not only enhance the accuracy of forecasts but also boost the robustness and adaptability of financial predictions, offering substantial advantages in environments that require real-time decision-making. The paper concludes with insights into the practical implications of neural forecasting in financial markets and recommendations for future research directions.

Complexity vs Empirical Score

  • Math Complexity: 5.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper includes advanced neural network architectures and mathematical formulations like basis expansion and time series decomposition, showing moderate-to-high mathematical density. It demonstrates empirical rigor by using real financial data from Yahoo Finance, implementing models in Python with specific libraries, and reporting comprehensive error metrics (MAE, RMSE, MAPE, SMAPE) for backtest-ready evaluation.

Research Flowchart

  flowchart TD
  A["Research Goal: Compare Neural Models N-HiTS & N-BEATS vs. Traditional Models for Financial Forecasting"] --> B{"Data Input: Financial Time Series Data"}
  B --> C["Preprocessing & Splitting"]
  C --> D["Model Training"]
  subgraph D ["Computational Processes"]
      D1["N-HiTS Model"]
      D2["N-BEATS Model"]
      D3["Traditional Statistical Models"]
  end
  D --> E["Forecasting & Evaluation"]
  E --> F{"Key Findings & Outcomes"}
  F --> G["Neural Models Superior Accuracy"]
  F --> H["Enhanced Robustness & Adaptability"]
  F --> I["Better Handling of Non-Linear Dynamics"]

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 comparative analysis of two neural forecasting models, N‑HiTS and N‑BEATS, for financial market prediction.

Source sentence
This research paper explores the effectiveness of two specific neural forecasting models, N-HiTS and N-BEATS, in predicting financial market trends.

Assumes 1

  • Historical financial market data consists of daily closing prices, volume, and other metrics for selected stocks and indices over the past decade. data
    Source sentence
    Data Sources Historical financial market data was collected from Yahoo Finance. This included daily closing prices, volume, and other relevant financial metrics for selected stocks and indices over the past decade

Data

  • Yahoo Finance historical market data · Yahoo Finance market-prices public
    past decade · selected stocks and indices
    Source sentence
    Data Sources Historical financial market data was collected from Yahoo Finance. This included daily closing prices, volume, and other relevant financial metrics for selected stocks and indices over the past decade

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • N‑BEATS consistently outperforms the other models across all error metrics (MAE, MSE, RMSE, MAPE, SMAPE). performance
    Source sentence
    The table demonstrates that NBEATS consistently out-performs the other models across all metrics, underscoring its effectiveness in financial forecasting applications.
  • Both N‑HiTS and N‑BEATS follow the trend of the actual data more closely than ARIMA, especially during periods of rapid change. predictive-accuracy
    Source sentence
    The figure demonstrates that both NHITS and NBEATS models follow the trend of the actual data more closely than the ARIMA model, especially during periods of rapid change
  • Figure 10 shows the close alignment of N‑BEATS with the testing data, suggesting superior forecasting accuracy compared to the other models. performance
    Source sentence
    Figure 10 presents the performance of the training, testing, and the forecasts from NHITS, ARIMA, and NBEATS models over the observed period. The visualization highlights the close alignment of NBEATS with the testing data, suggesting superior forecasting accuracy compared to the other models.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 28 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive (1 by independent authors)

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.