N ew Approaches to Forecasting

It is the data in the GPS of your phone, your searches on the web, the friends in your network, the things you purchase, etc. [...] Lack of knowledge about the pre-treatment of the raw data, along with a lack of access to the same, introduces some opacity about the measurement quality of the data (Kapetanios and Papailias, 2018). [...] Thus, the latter will have to be aggregated to match the frequency of the target variable: offsetting the gains that may have been derived from the higher frequency and volume of the big data being used (Buono et. [...] 15 to efficiently find the direction and magnitude of change needed in the weights and biases used in all the steps of the second and subsequent iterations12. [...] Once the basic pre-checks and necessary data transformations have been conducted, the forecaster needs to select the relevant predictors from the set of the available n, also known as ‘feature selection’ in the machine learning literature.