Add logic to `DefaultAlgorithm` to run baseline naive pipelines with and without engineered features

This issue tracks adding logic to `DefaultAlgorithm` to run naive pipelines with and without engineered features. This will allow us to creates heuristics on where or not to use selected engineered featuers.

```
def _create_naive_pipelines_with_raw_features(self):
engineered_features = self.get_engineered_features()
estimators = self._naive_estimators()
parameters = self._pipeline_params if self._pipeline_params else None
parameters.update('Drop Columns Transformer': {'columns': engineered_features})

pipelines = [
make_pipeline(
self.X,
self.y,
estimator,
self.problem_type,
sampler_name=self.sampler_name,
parameters=parameters,
extra_components=DropColumns,
)
for estimator in estimators
]

pipelines = self._create_pipelines_with_params(pipelines, parameters={})
return pipelines
```