we have just released the last part of additional data for this challenge. This data set describes individual events logged by security systems, that correspond to investigated alerts from the training and test sets. As in the case of localized alerts data, you can match the records from the new table with records from training and test sets using identifiers from the ' alert_id' column.
Any team that achieves the preliminary score higher than 0.8500 should be able to see the new data in the Data files section of the page.
Please be aware that additional data table is quite large - its total size after decompression is approximately 430GB. For this reason, we decided to divide this table into 29 smaller chunks. Each chunk, after unzipping, has between 10GB and 20GB. For convenience, data regarding any single investigated alert (a record from the original data) is always stored in a single chunk.
The event logs are missing for approximately 10% of data.
Good luck and have fun :-)
I am curious if you can get a better score with newly released data. Did you manage to download the files?
If you have any problems or questions, feel free to ask.
Good luck with new data