Add xgboost example using Bayesian optimization

[richardsliu]

This example uses the image built from https://github.com/kubeflow/examples/tree/master/xgboost_ames_housing.

The hyperparameters being tuned are:

• Number of boost trees
• learning rate

We are trying to minimize the mean_absolute_error.

---

This change is 

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: richardsliu

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [richardsliu]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[richardsliu]

/hold

[johnugeorge]

Is it using default metric collector?

[richardsliu]

@johnugeorge Yes, the default metric collector works here (assuming that the training worker outputs logs in the expected format).

[hougangliu]

Reviewable status: 0 of 1 files reviewed, 2 unresolved discussions (waiting on @richardsliu, @ddysher, and @texasmichelle)

---

examples/xgboost-bayesian-example.yaml, line 17 at r1 (raw file):

  requestcount: 10
  metricsnames:
    - mean_absolute_error

mean_absolute_error is not in need here since it is objectivevaluename

---

examples/xgboost-bayesian-example.yaml, line 61 at r1 (raw file):

              - name: datadir
                persistentVolumeClaim:
                claimName: claim

please add pvc yaml in this patch, too

[hougangliu]

Reviewable status: 0 of 1 files reviewed, 3 unresolved discussions (waiting on @richardsliu, @ddysher, and @texasmichelle)

---

examples/xgboost-bayesian-example.yaml, line 13 at r1 (raw file):

  owner: crd
  optimizationtype: minimize
  objectivevaluename: mean_absolute_error

I wonder if https://github.com/kubeflow/examples/blob/master/xgboost_ames_housing/housing.py can print mean_absolute_error in log

[richardsliu]

@hougangliu The example docker image is being fixed in this PR: kubeflow/examples#476. I added comments clarifying the prerequisites.

[hougangliu]

/lgtm