Prediction Models

Our prediction models allow you to create quantitative models using advanced statistical methods, through a relatively simple interface. We aim to make the job on your part as simple as possible, while the heavy lifting is done behind the scenes by our modelling engine.

📘

Prediction models in a nutshell

Before starting modelling, you should have in mind the purpose of the modelling. You pick the data you want to use and tell us which relationships you want to explore or which time series you want to predict, and we do the rest.

With KPI modelling, we are trying to predict a number ahead of its report date. It can be a fundamental value the company reports, a factor the data vendors report/aggregate on a monthly basis or just the next 1 month stock price returns we think a company would have based on certain input metrics we have at our hand. We can choose to predict the absolute value, a range through a classification model or just the direction for change in an underlying metric.