Improve unit test temp file usage

[woodsp-ibm]

Summary

As it was two separate unit tests created/used the same named file in the temp directory. Running separately things are fine and are most often in CI - but CI runs tests in parallel and this has led to occasional failures for example in this log with the relevant excerpt as below

==============================
Failed 2 tests - output below:
==============================

test.algorithms.regressors.test_qsvr.TestQSVR.test_save_load
------------------------------------------------------------

Captured traceback:
~~~~~~~~~~~~~~~~~~~
    Traceback (most recent call last):

      File "/Users/runner/work/qiskit-machine-learning/qiskit-machine-learning/test/algorithms/regressors/test_qsvr.py", line 123, in test_save_load
    FakeModel.load(file_name)

      File "/Users/runner/work/qiskit-machine-learning/qiskit-machine-learning/qiskit_machine_learning/algorithms/serializable_model.py", line 54, in load
    model = dill.load(handler)
            ^^^^^^^^^^^^^^^^^^

      File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/dill/_dill.py", line 287, in load
    return Unpickler(file, ignore=ignore, **kwds).load()
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

      File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/dill/_dill.py", line 442, in load
    obj = StockUnpickler.load(self)
          ^^^^^^^^^^^^^^^^^^^^^^^^^

    EOFError: Ran out of input


test.algorithms.regressors.test_fidelity_quantum_kernel_qsvr.TestQSVR.test_save_load
------------------------------------------------------------------------------------

Captured traceback:
~~~~~~~~~~~~~~~~~~~
    Traceback (most recent call last):

      File "/Users/runner/work/qiskit-machine-learning/qiskit-machine-learning/test/algorithms/regressors/test_fidelity_quantum_kernel_qsvr.py", line 112, in test_save_load
    regressor_load = QSVR.load(file_name)
                     ^^^^^^^^^^^^^^^^^^^^

      File "/Users/runner/work/qiskit-machine-learning/qiskit-machine-learning/qiskit_machine_learning/algorithms/serializable_model.py", line 53, in load
    with open(file_name, "rb") as handler:
         ^^^^^^^^^^^^^^^^^^^^^

    FileNotFoundError: [Errno 2] No such file or directory: '/var/folders/24/8k48jl6d249_n_qfxwsl6xvm0000gn/T/qsvr.model'

    
During handling of the above exception, another exception occurred:


    Traceback (most recent call last):

      File "/Users/runner/work/qiskit-machine-learning/qiskit-machine-learning/test/algorithms/regressors/test_fidelity_quantum_kernel_qsvr.py", line 127, in test_save_load
    os.remove(file_name)

    FileNotFoundError: [Errno 2] No such file or directory: '/var/folders/24/8k48jl6d249_n_qfxwsl6xvm0000gn/T/qsvr.model'

Details and comments

This changes from specific named temp file to creating a temp folder with that named file in it, same as is done in another similar test here

qiskit-machine-learning/test/algorithms/regressors/test_neural_network_regressor.py

Line 187 in a0ea90a

with tempfile.TemporaryDirectory() as dir_name:

where the temp folder is automatically removed, along with the contents, when the context ends.

This change should should eliminate that occasional random CI failures where the two tests are run at the same time and there was collision in the file used. By using a temp folder each test will have their own folder and any collision avoided from use of the same named file since they will be in different folders. This also allows multiple instances of the same test to be run in parallel, not that CI does this, but its a more general fix.

[adekusar-drl]

Thanks for taking care of this!

[coveralls]

Pull Request Test Coverage Report for Build 6421477684

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 92.689%

---

Totals
Change from base Build 6419328274:	0.0%
Covered Lines:	1851
Relevant Lines:	1997

---

💛 - Coveralls