AutoML.Net improvements tracking down list

[LittleLittleCloud]

• AutoML Sweepable API (Proposal: Sweepable API #5993)

  • implementation (Use SweepablePipeline #6285)
  • make it public (make AutoMLExperiment public && some small refactor #6173)
  • documents
    • sweepable pipeline
  • code example
  • jupyter notebook example
    • forecast using ssa
    • automl sweepable api

• AutoML Featurizer API (add auto featurizer api #6187)

  • implementation (implement auto featurizer #6205) (add image featurizer to AutoFeaturizer #6261)
  • make it public
  • documents
  • code example
    • image classification with image featurizer
  • jupyter notebook example
    • regression with taxi fare
    • classification with titanic

• AutoML Experiment API (Proposal: Experiment API #6118 )

  • implementation
  • make it pubic (make AutoMLExperiment public && some small refactor #6173)
  • documents [Create example for AutoMLExperiment in api doc. #6589]
  • code example
  • jupyter notebook example
    • titanic
    • taxi-fare
    • iris
    • luna forecasting using regression
    • needs more
  • reimplement existing experiment using AutoML Experiment API
    • BinaryExperiment (reimplement binary experiment using AutoMLExperiment #6246)
    • Multiclass Experiment (improve multiclassification using AutoMLExperiment #6270)
    • Regression (Improve RegressionExpeirment using AutoMLExperiment #6338)
    • Recommendation
    • Ranking

• AutoML Advance Feature

  • continue training [Add ITrialResultManager for continue training in AutoML #6335]
  • exit strategy [Add MaxModelToExplore exit strategy to AutoMLExperiment. #6402 ]
  • limit resource (CPU/memory) [Add SetMaximumMemoryUsageInMegaByte in AutoMLExperiment #6305]

[michaelgsharp]

@LittleLittleCloud I'll tag this for ML.NET 2.0 (For release this November) just so it can be tracked appropriately. If the plan is not to have it by Nov let me know And I can push it to a later milestone.

[luisquintanilla]

@LittleLittleCloud it looks like most of the items here are complete. Are we okay to close? Also I'm tracking 3.0 improvements in this issue.

#6514

We can maybe migrate anything outstanding there.