Chapter 3: Understanding How to Predict Industry-Wide Trends Using Big Data
Forecasting is a tricky business; no one is sure why it is that some forecasts are right and others are wrong, but two main factors contribute to forecast accuracy:
- Which data and models are used
- What assumptions are made about the variables being forecasted
Unfortunately, as this chapter will show, most traditional methods of forecasting suffer from low predictive accuracy because they do not take these important factors into account properly. Here, we will explore how big data changes all that by enabling better predictions.
Our goal is not to present you with yet another prediction tool (though we are going to discuss a few). Instead, we want to share some insights about why conventional methods fail and how we might harness big data to make better predictions ourselves. These insights
will set product teams up for success since accurate forecasts of market demand and sentiment...
