In this notebook I'm giving away a strategy that will likely give you an edge if implemented correctly. In short, it might be possible to reverse engineer the injected CW signal by using raw Ligo&Virgo data.
\nTL;DR
\nThis is clearly not the solution organizers had in mind, but it's completely legitimate according to the rules: https://www.kaggle.com/competitions/g2net-detecting-continuous-gravitational-waves/rules
", "messages": [ { "id": 2042667, "postDate": "2022-11-24T21:23:07.407Z", "content": "In this notebook I'm giving away a strategy that will likely give you an edge if implemented correctly. In short, it might be possible to reverse engineer the injected CW signal by using raw Ligo&Virgo data.
\nTL;DR
\nThis is clearly not the solution organizers had in mind, but it's completely legitimate according to the rules: https://www.kaggle.com/competitions/g2net-detecting-continuous-gravitational-waves/rules
", "rawMarkdown": "[In this notebook](https://www.kaggle.com/vslaykovsky/g2net-winning-strategy-with-external-data) I'm giving away a strategy that will likely give you an edge if implemented correctly. In short, it might be possible to reverse engineer the injected CW signal by using raw Ligo&Virgo data.\n\n**TL;DR**\n\n1. Get raw detector data at https://www.gw-openscience.org/archive/O3a_4KHZ_R1/\n2. Generate SFTs using instructions at https://youtu.be/A9iWRcmG0Rs?t=2199 (you'll likely need to google for more details on this step). Use GPS timestamps from test set to produce SFTs.\n3. Read generated SFTs using pyfstat.utils.sft.get_sft_as_arrays\n4. Match generated SFTs with real samples from test set.\n5. Find the difference between SFTs and real samples. difference > EPS -> label==1; difference <= EPS -> label==0.\n\nThis is clearly not the solution organizers had in mind, but it's completely legitimate according to the rules: https://www.kaggle.com/competitions/g2net-detecting-continuous-gravitational-waves/rules \n\n", "votes": 18 } ], "comments": [], "raw_markdown_by_id": { "2042667": "[In this notebook](https://www.kaggle.com/vslaykovsky/g2net-winning-strategy-with-external-data) I'm giving away a strategy that will likely give you an edge if implemented correctly. In short, it might be possible to reverse engineer the injected CW signal by using raw Ligo&Virgo data.\n\n**TL;DR**\n\n1. Get raw detector data at https://www.gw-openscience.org/archive/O3a_4KHZ_R1/\n2. Generate SFTs using instructions at https://youtu.be/A9iWRcmG0Rs?t=2199 (you'll likely need to google for more details on this step). Use GPS timestamps from test set to produce SFTs.\n3. Read generated SFTs using pyfstat.utils.sft.get_sft_as_arrays\n4. Match generated SFTs with real samples from test set.\n5. Find the difference between SFTs and real samples. difference > EPS -> label==1; difference <= EPS -> label==0.\n\nThis is clearly not the solution organizers had in mind, but it's completely legitimate according to the rules: https://www.kaggle.com/competitions/g2net-detecting-continuous-gravitational-waves/rules \n\n" } }