Due to the nonlinearity, high volatility, and multi factor driving characteristics of financial time series, the fitting ability of traditional prediction models is limited. This study focuses on financial time series prediction and multi-scale correlation analysis by integrating network big data and deep learning. The MF-DCCA method is used to verify the correlation between the RMB exchange rate and Baidu Index, WTI oil price and Google Trends. A WOA-STL-LSTM optimization model is constructed by integrating network search data as external features, significantly improving the accuracy of exchange rate and oil price prediction. Study the use of Hurst exponent, fractal dimension, Lyapunov exponent, and transfer entropy to quantify data complexity, revealing the marginal effect and keyword difference mechanism of Google Trends on WTI prediction. To address the issue of small sample size, fractal interpolation and linear interpolation are used to enhance data granularity. The adaptation strategy is validated by combining LSTM, GRU, and Bi LSTM models. It is found that fractal interpolation is superior to linear interpolation in simulating complex fluctuations, and the model selection needs to match data characteristics - Bi LSTM/LSTM is better in small sample scenarios, and the combination of GRU and fractal interpolation, Bi LSTM and linear interpolation after data augmentation yields the best results. Research provides an interpretable framework for financial time series forecasting, promoting the development towards precision and transparency.

Jingzhi Yin. Research on Financial Time Series Prediction and Multiscale Correlation Based on the Fusion of Network Big Data and Deep Learning. International Journal of Big Data Intelligent Technology (2026), Vol. 7, Issue 1: 52-59. https://doi.org/10.38007/IJBDIT.2026.070107

[1] Yuan J, Li J, Hao J. A reliable ensemble forecasting modeling approach for complex time series with distributionally robust optimization. Computers and Operations Research, 2025, 173.

[2] Okwonu F Z, Chiyeaka O M, Ahad N A, et al. ROBUST PEARSON CORRELATION COEFFICIENT FOR IMBALANCED SAMPLE SIZE AND HIGH DIMENSIONAL DATA SET. Science World Journal, 2025, 20(1).

[3] Zheng L, Chai H, Chen X, et al. Search-based Time-aware Graph-enhanced Recommendation with Sequential Behavior Data. ACM Transactions on Recommender Systems, 2024, 2(4).

[4] Gharehbaghi A, Ghasemlounia R, Ahmadi F, et al. Developing a novel hybrid model based on GRU deep neural network and Whale optimization algorithm for precise forecasting of river's streamflow. Scientific Reports, 2025, 15(1).

[5] Xindi Wei. Optimization of Machine Learning Models and Application Supported by Data Engineering. Machine Learning Theory and Practice (2025), Vol. 5, Issue 1: 117-124

[6] Yiting Gu. The Strategic Application of Front-End Technology in The Process of Digital Transformation. Machine Learning Theory and Practice (2025), Vol. 5, Issue 1: 125-132.

[7] Huijie Pan. Design of Data-Driven Social Network Platforms and Optimization of Big Data Analysis. Machine Learning Theory and Practice (2025), Vol. 5, Issue 1: 133-140.

[8] Yixian Jiang. Research on Integration and Optimization Strategies of Cross-platform Machine Learning Services. Machine Learning Theory and Practice (2025), Vol. 5, Issue 1: 141-148.

[9] Shuang Yuan. Integration and Optimization of Network Security Protection Strategies and Vulnerability Detection Technologies. International Journal of Neural Network (2025), Vol. 4, Issue 1: 32-39.

[10] Jiangnan Huang. Application of AI-driven Personalized Recommendation Technology in E-commerce. International Journal of Neural Network (2025), Vol. 4, Issue 1: 40-47.