Machine Learning: Unlocking the Power of Unstructured Data
Establishing a link between current observables — structured and unstructured data — and future behavior is a major undertaking in finance, and naturally the field is embracing machine learning to tackle this typical learning task. This requires broadening the range of data and methods, and shaking some old habits.
Conventionally, the majority of financial data was processed in structured form — such as numerical information from markets and security prices — and the methods to process it were borrowed from the standard statistical toolkit. Advances in machine learning and processing power mean it is now possible to process (and make sense of) vast amounts of unstructured data, which brings with it the potential to transform the industry.