Predicting Fine-grained Regional Economic data based on multiple data sources
Details
| Supervisor: | Shichang Ding,Please contact sding@gwdg.de or scdingwork@gmail.com |
| Duration: | 3 MONTH |
| Type: | Student Project (plus potential Master's thesis) |
| Status: | open |
Title: Predicting Fine-grained Regional Economic data based on multiple data sources （基于多数据源的细粒度地区经济数据预测）
What to do: (take part in) designing and implementing a system for predicting regional economic status based on multiple data sources
Why important: Regional economic data(e.g, GDP and per disposable income) is important to both governments and companies. However, the manual survey costs lots of time and money. And in many countries, the data is only public on city-level or county-level. Predicting Fine-grained Regional Economic data based on multiple data sources like Smart card data, house prices or Online job websites would be of great help.
Requirements: Python, basic knowledge about Semi-supervised learning or Web Crawling
Do not worry if you do not have too much knowledge in this topic. At least, you can try to implement a web crawler at first, and then we can show you how to design an end-end DM system. However, if you are interested in semi-supervised learning, you can directly take part in the system designing and implementing.
WHY you should pick this project: Both Web crawling and semi-supervised learning are widely-used techniques of many companies. And this topic is really interesting and useful, considering no companies want to do a large-scale manual survey.
Further task: Because an end-end system usually needs a lot of time to complete, temporarily there is no plan for further task.
About Deep Learning (DL): Recently, there are several new DL methods about semi-supervised learning. However, none of them have been tested on our dataset yet. Any try will be welcome. However I have to warn you that there is no guarantee for success.
The work is predominantly targeted at an approximately 3-month student project, but can be extended into a Master's thesis as well.