[MRG+1] Add 'concurrency' block to GHA and fix build warnings

[aaronreidsmith]

Description

This PR makes it so if we push a commit to the same branch, it will cancel previous builds on GHA, saving us some cycles.

It also fixes a couple warnings in the build:

• np.int -> int
• np.bool -> bool
• np.float -> float

• LinearRegression(normalize=True) was apparently deprecated by sklearn in favor of make_pipeline(StandardScaler(with_mean=False), LinearRegression())

  /Users/runner/hostedtoolcache/Python/3.7.12/x64/lib/python3.7/site-packages/sklearn/linear_model/_base.py:145: FutureWarning: 'normalize' was deprecated in version 1.0 and will be removed in 1.2.
  If you wish to scale the data, use Pipeline with a StandardScaler in a preprocessing stage. To reproduce the previous behavior:
  
  from sklearn.pipeline import make_pipeline
  
  model = make_pipeline(StandardScaler(with_mean=False), LinearRegression())
  
  If you wish to pass a sample_weight parameter, you need to pass it as a fit parameter to each step of the pipeline as follows:
  
  kwargs = {s[0] + '__sample_weight': sample_weight for s in model.steps}
  model.fit(X, y, **kwargs)

Type of change

• [X] CI/CD change
• [X] Fix deprecations

How Has This Been Tested?

• [X] Tests still pass
• [X] Most warnings are gone
• [X] Old builds cancel on push to same branch

Checklist:

N/A

[aaronreidsmith]

Only remaining warnings are these, which I think is fine? They are on our side at least, and I think we are intentionally creating those warnings. Could be wrong

=============================== warnings summary ===============================
tests/test_pipeline.py::test_pipeline_predict_inverse_transform[True-True-X0-pipeline5]
tests/test_pipeline.py::test_pipeline_predict_inverse_transform[True-True-X1-pipeline5]
  /Users/runner/hostedtoolcache/Python/3.7.12/x[64](https://github.com/alkaline-ml/pmdarima/runs/6054174506?check_suite_focus=true#step:11:64)/lib/python3.7/site-packages/pmdarima/preprocessing/endog/boxcox.py:161: RuntimeWarning: overflow encountered in exp
    return np.exp(y) - lam2, exog

tests/test_pipeline.py::test_pipeline_predict_inverse_transform[True-True-X0-pipeline6]
tests/test_pipeline.py::test_pipeline_predict_inverse_transform[True-True-X1-pipeline6]
  /Users/runner/hostedtoolcache/Python/3.7.12/x64/lib/python3.7/site-packages/pmdarima/preprocessing/endog/boxcox.py:165: RuntimeWarning: invalid value encountered in power
    de_exp = numer ** (1. / lam1)  # de-exponentiate