MIT EECS | Hudson River Trading Undergraduate Research and Innovation Scholar
Quantum Time Series Forecasting
Dirk R. Englund
A time series is any dataset that captures n-dimensional feature vectors varying over time. Time series forecasting deals with making future predictions based on existing time series data. Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA) and Autoregressive Integrated Moving Average (ARIMA) are models commonly used classically. As the dimension of the feature vectors grows, these algorithms become increasingly difficult to efficiently implement to produce accurate results. Quantum computing is motivating the development of algorithms that are exponentially faster than their classical counterparts, and could provide speedup or improve the accuracy of existing classical forecasting algorithms. Our goal is to quantize time series forecasting and determine the potential advantages quantum computing could offer in this field.
I absolutely love research, particularly theory. I find thinking about how quantum computing can be applied to different fields really exciting! Quantum computing is a very new field so there is a lot of room for innovation. I chose to participate in SuperUROP because I wanted to a chance to take a deep dive into this work, and receive training on how to conduct research effectively.