Data/ML/AI
Time Series Forecasting
Also written as Demand Forecasting, Forecasting Model, Seasonality
Predicting future values of something measured over time — sales, load, prices — where the order of the data matters and an ordinary random shuffle of the training data quietly cheats.
Think of it like
Forecasting weather rather than sorting photographs. Yesterday is evidence about tomorrow in a way one photo is not about another.
Junior or senior?
A common home for subtle, invisible mistakes.
Senior sounds like
Knows why they had to validate on later time periods instead of a random split.
Ask them
“How did you split training and test data, given the data was ordered in time?”