Survey & development of new methods for dynamic hierarchical time series models
Description
The objective here is to acquaint students with a real world problem and address it through an understanding of the state of the art in hierarchical time series with a dynamic component. After a throrough review reflected in a technical survey, students will proceed to implement a method of their choice or develop a new one and assess it against the state of the art on a real-world problem
Objective
- Development of new methods for dynamic hierarchical time series
- Polishing of software development skills in python
- Understanding of the state of the art in time series modeling
Data sets
- Private data set will be shared individually
- https://www.kaggle.com/c/walmart-recruiting-store-sales-forecasting
- https://www.dunnhumby.com/sourcefiles
Target audience
Ideal for students interested in developing new methods for time series with a dynamic hierarchy