Forecast: Principle and Practice

Chapter-1 

Exercise

Vehicle resale value forecast 

f(Number of kms, accident history, model year, Is model in sale right now by company,  lease duration, advertising campaigns, errors)

Weekly air passenger traffic distribution

f(weekly traffic data, school holidays, major sporting events, advertising campaigns, competition behaviour, error)

Five steps

Define the problem

Need to forces vehicle resale value. IT should include lease duration so that user can play around for adjusting lease policy in order to maximise profit

It should include advertising campaigns so that it will help sales team to adjust values

Short term forecast will help team for better control of profits

Collect data

Preliminary analysis

Can be done via scatter plot.. to see patterns in the profit as time series data

Create the model

Validate the model