Download fracture_identification folder from the git repo (type in "git lfs clone https://github.com/pddasig/Machine-Learning-Competition-2024.git" in your command prompt) and put it into your google drive (/content/gdrive/My Drive/fracture_identification/). Then you can run a draft notebook "2024_PDDA_ML_competition.ipynb" for fracture detections. If you have any questions or need clarifications, send an email to "pdda_sig@spwla.org". Thank you!
Note: We will update this README for answering questions and sharing information, stay tuned!
UPDATES: By submitting your work to this contest, you grant the organizers permission to publish your project in a peer-reviewed academic journal. This is a condition of entry and participation. The goal is to promote knowledge sharing and contribute to the broader machine learning community.
This agreement applies to all submissions, regardless of whether they win. Each participant must ensure their work is original and does not infringe on any intellectual property rights. Further details about the publication process will be provided to all eligible participants after the contest concludes.
We look forward to seeing your innovative work!
UPDATES:
- For the submission, it seems like the Codalab bug is caused by the deprecation of mutation events in Chrome (probably Edge and Explorer too). It still runs on Firefox so participants are recommended to submit using Firefox.
- Competition is successfully completed and the answers for fracture locations and their images are uploaded in our repo.
- We are designing our next ML competition and you are welcome to submit your idea or suggestions! (pdda_sig@spwla.org)
===========================================================
Announcing the Winners of SPWLA PDDA 2024 ML Competition!
We are thrilled to announce the winners of the SPWLA Petrophysical Data-Driven Analysis (PDDA) 2024 Machine Learning Competition! This year’s competition saw outstanding participation from talented teams worldwide, pushing the boundaries of resistivity image log analysis for fracture identifications through innovative machine learning approaches.
Top 5 Winning Teams: 🥇 Vdehdari – F1 Score: 0.6667 🥈 Atwah_Analytics – F1 Score: 0.3200 🥉 We_will_log_you – F1 Score: 0.2449 🏅 SWPU_WELL_LOGGING_TOP1 – F1 Score: 0.2222 🏅 PETAI – F1 Score: 0.2128
Each of these exceptional teams demonstrated cutting-edge methodologies and exceptional predictive accuracy in test well evaluations. As part of their achievement, the winning teams will receive awards and will be co-authors of an upcoming Petrophysics Journal paper, highlighting their innovative contributions to the field.
We extend our sincere gratitude to all participants, judges, sponsors, and organizers for making this competition a huge success. Stay tuned for further details on upcoming opportunities and advancements in petrophysical machine learning!
Congratulations once again to the winners!
SPWLA PDDA 2024 ML Competition Committee
