Develop a forecasting model

Problem description

"Click count" refers to the number of visits/sales of products on a particular date. Based on the data, we have visible annual seasonality with few surges of demand across the year. The organization intends to develop a method using machine learning capabilities to get a look ahead on future sales. The primary concern is to have a forecast with high accuracy during the peak season without compromising the solution quality for rest of the year. Based on the forecast, inventory planning and necessary investments will be determined.

Expectations

Exploratory data analysis to obtain useful insights

Development of features that enhances ML model performance

Implementation of one or more machine learning models to obtain good results

Ability to forecast at least 30 days into the future i.e. last date in the problem + 30 days

Reasonably good model performance i.e. MAPE 10% or less, including peak season

Статистика Алгоритмы Python Machine Learning (ML) Deep Learning

ID проекта: #33547055

О проекте