Time Series Prediction

October 2020

London, United Kingdom

In response to a data science challenge issued by causaLens, I created a new time series dataset covering the period between the 2nd of March and the 21st of September, 2020. I proposed using the dataset to predict the proportion of residents in the Greater London area who express boredom in a given week. I then developed two models for the problem based respectively on an auto-regressive moving average with exogenous inputs and a gated recurrent unit.