Adjust probability estimation function used in linear tree-based method

[tic66777]

What does this PR do?

Because the probability estimation function of tree-based method is not appropriate, this PR implements a tunable sigmoid function that can give better probability estimation.

Test CLI & API (bash tests/autotest.sh)

Test APIs used by main.py.

• Test Pass
  • (Copy and paste the last outputted line here.)
• Not Applicable (i.e., the PR does not include API changes.)

Check API Document

If any new APIs are added, please check if the description of the APIs is added to API document.

• API document is updated (linear, nn)
• Not Applicable (i.e., the PR does not include API changes.)

Test quickstart & API (bash tests/docs/test_changed_document.sh)

If any APIs in quickstarts or tutorials are modified, please run this test to check if the current examples can run correctly after the modified APIs are released.

[chcwww]

I came up with an idea that we could mimic how we handled LINEAR_TECHNIQUES by using a dict like

ESTIMATORS = {
    "exp-L1": lambda x, _: np.maximum(0, 1 - x),
    "exp-L2": lambda x, _: np.square(np.maximum(0, 1 - x)),
    "sigmoid-A": lambda x, param: -log_expit(param * x),
}

in the linear/utils.py.
And implement the estimator using a functor like this:

class Estimator:
    def __init__(self, estimation_function: str, estimation_parameter: int = None):
        self.estimation_function = ESTIMATORS[estimation_function]
        self.estimation_parameter = estimation_parameter

    def __call__(self, x):
        return self.estimation_function(x, self.estimation_parameter)

Would that be cool?