wty9391 / bid-lands

Functional Bid Landscape Forecasting for Display Advertising

This is a repository of the experiment code supporting the paper Functional Bid Landscape Forecasting for Display Advertising in ECML 2016.

Demo

After pulling the repository, you could start from checking the demo under the folder of python by running:

$ python demo.py

After running, you could get the performance table printed in the console like:

          ANLP                                    KLD
Campaign  MM      NM      SM      NTM     STM     MM      NM      SM      NTM     STM
2259      7.2340  6.3751  5.6045  5.4486  5.0781  0.9165  0.9245  0.2770  0.9024  0.1806
overall   7.2340  6.3751  5.6045  5.4486  5.0781  0.9165  0.9245  0.2770  0.9024  0.1806

Note these results are produced from only a small part of the data (the first 300,000 lines for campaign 2259 in iPinYou).

Large-scale Experiment

For the large-scale experiment, please first check the GitHub project make-ipinyou-data for pre-processing the iPinYou data. After downloading the dataset, by simplying make all you can generate the standardised data. Please merge the make-ipinyou-data folder under /bid-lands. You could run the experiment by running:

$ python main.py