This is an old revision of the document!


Forecasting Meeting Minutes for Week of December 6, 2016

Present Tuesday: Gordon, Brieanna, Jaimie

  • Check Accuracy
    • RMSE: 200
    • Persistence: Use last data point as estimate
      • Current time = prediction
      • 1 tap = 1
      • No bias
      • No normalization
    • Create estimate: All data and guess mean
    • Max solar irradiance is 1000 W
    • Check with Travis's Report
  • Test many data sets
    • Test RSME should be higher than train RMSE
    • Varies on the number of taps
  • forecasting/meeting_minutes_december_6_2016.1481079123.txt.gz
  • Last modified: 2021/09/19 21:59
  • (external edit)