Add noise

[joewandy]

• Background noise.
  • We need to extract from the real data:
  • Peaks in regions of interest
  • Non-peaks in regions of interest
  • Non-peaks not in regions of interest
• Small peaks.
• Noise to rt, intensities, mz values
  • Don’t add noise to empirical data, except for rt shift across samples.
• Adversarial approach?

[joewandy]

Not sure what @sdrogers means by the 'adversarial approach' above.