Skip to content

Latest commit

 

History

History
44 lines (37 loc) · 1.84 KB

progress.md

File metadata and controls

44 lines (37 loc) · 1.84 KB

Progress in the gravitational ocean

25-Oct-2022

  • understand the challenge
  • see PPP

01-Nov-2022

  • setup repo
  • first tests with efficentnet

08-Nov-2022

  • Experimented with fastAI
  • play with signal-to-noise-ratio:
    • high ratio: good signals -> AI learns and predicts well
    • low ratio: little signal -> not get better than guessing (0.5)
  • problems with Charlie
  • looked at papers (paper1, paper2)
  • started visualizing data

15-Nov-2022

  • first submission: cheated the challenge, no AI, just interpolation (see cheat-task notebook kaggle)
  • started crnn from this link
  • created new dataset from kaggle training data
  • our generated data is not really suitable for training, because it has 5000 frequencies, but real data only has 360

22-Nov-2022

  • fixed problems with CRNN, works now
  • set fixed sequence length for crnn, it does not need to guess
  • use tensorboard to visualize training progress
  • read this paper -> increase kernel size
  • setup data and code on charlie -> took a while

29-Nov-2022

  • analyzed kaggle solution discussed last week
  • adding more noise is valid augmentation (maybe even the best augemtation)
  • setup and successfully develop stuff on charlie (using ssh, git repo)

08-Dec-2022

  • prepared presentation for last week Thursday
  • started with transformer from this article
  • understood template code and modified to work with the trained CNN data
  • adjusted dataloader
  • train 2/3 different transformers on CNN output (33 h of training)
  • with transformer and seq_len 16 we are way better than guessing, accuracy around 62 %