From the course: Problem Identification and Solution Design for Data Scientists

Unlock the full course today

Join today to access over 24,800 courses taught by industry experts.

Prediction vs. forecasting

Prediction vs. forecasting

From the course: Problem Identification and Solution Design for Data Scientists

Start my 1-month free trial Buy for my team

Prediction vs. forecasting

“

- [Presenter] Most clients and project sponsors think they want to use the latest machine learning and AI methods to do forecasting, but in my experience, presented with the option, predictive analytics is often the better fit. Forecasting, for instance, like sales or revenue forecasting might sound like a kind of prediction and it certainly has a predictive element to it, but it's much more like inference. The reason is, unlike the kind of predictive analytics that we've been discussing, forecasting is about an overall pattern and not about individual cases. Also, sometimes, they're asking you if you can do something with AI to do a forecast because they're not satisfied with their current forecast. It's often the job of the finance team and not the data science team because training and time series forecasting is typically part of a finance degree, so watch out for this scenario. It potentially puts you in competition with another team, which might be okay if it's fully transparent,…

Contents