Financial Time Series Data Processing for Machine Learning
This work addresses data preprocessing challenges for financial analysts using machine learning, but it is incremental as it builds on existing methods without introducing major innovations.
The paper tackles the problem of preprocessing financial time series data for machine learning by comparing scaling methods for stationarity and information preservation, and proposes empirical tests and data split techniques to avoid overfitting.
This article studies the financial time series data processing for machine learning. It introduces the most frequent scaling methods, then compares the resulting stationarity and preservation of useful information for trend forecasting. It proposes an empirical test based on the capability to learn simple data relationship with simple models. It also speaks about the data split method specific to time series, avoiding unwanted overfitting and proposes various labelling for classification and regression.