One Hot Encoder: Drop one redundant feature by default for features with two categories

Our one hot encoder creates a feature for every level of the original categorical feature:

from evalml.pipelines import OneHotEncoder
import pandas as pd
df = pd.DataFrame({"category": ["a", "b"], "number": [4,5 ]})
OneHotEncoder().fit_transform(df).to_dataframe()

The category_a and category_b columns are completely collinear, which makes one redundant. This could have adverse effects on estimator fitting. I think we should drop one by default.

FYI @rpeck

[rpeck]

Third Law of Code: Thou Shalt Not Make == Comparisons With Floats

[angela97lin]

Post-discussion with @freddyaboulton @rpeck @dsherry @chukarsten @jeremyliweishih

• We will only do this for binary cases.
• A “nice-to-have” is to use, in the binary case, is the minority class, but otherwise just choosing one of the two categories should suffice.