Transformers with Attentive Federated Aggregation for Time Series Stock Forecasting
This work addresses privacy and data scarcity issues in time series forecasting for enterprises, but it is incremental as it builds on existing transformer and federated learning techniques.
The paper tackled the problem of adapting transformers to time series stock forecasting by proposing an attentive federated aggregation method, which achieved better performance while preserving privacy, as shown in empirical results on Yahoo! Finance stock data.
Recent innovations in transformers have shown their superior performance in natural language processing (NLP) and computer vision (CV). The ability to capture long-range dependencies and interactions in sequential data has also triggered a great interest in time series modeling, leading to the widespread use of transformers in many time series applications. However, being the most common and crucial application, the adaptation of transformers to time series forecasting has remained limited, with both promising and inconsistent results. In contrast to the challenges in NLP and CV, time series problems not only add the complexity of order or temporal dependence among input sequences but also consider trend, level, and seasonality information that much of this data is valuable for decision making. The conventional training scheme has shown deficiencies regarding model overfitting, data scarcity, and privacy issues when working with transformers for a forecasting task. In this work, we propose attentive federated transformers for time series stock forecasting with better performance while preserving the privacy of participating enterprises. Empirical results on various stock data from the Yahoo! Finance website indicate the superiority of our proposed scheme in dealing with the above challenges and data heterogeneity in federated learning.