Orientalists are logistic regression and Westerners are decision trees.
https://gyazo.com/65041f75b35b16a9903a8525291966c5
Should the target in the figure be in group 1 or group 2?
Orientals tend to choose Group 1 and Westerners Group 2.
Represent petal, inflorescence, leaf, and stem features as (1, 1, 1, 1) targets
Group 1
(1, 1, 1, 0)
(1, 0, 1, 0)
(1, 1, 0, 0)
(0, 1, 1, 0)
Group 2
(0, 0, 1, 1)
(0, 0, 0, 1)
(0, 1, 0, 1)
(1, 0, 0, 1)
Let us call the axes (a, b, c, d), respectively
Western identification discards information other than the d-axis to identify
Oriental-type identification uses all axes.
code:python
import numpy as np
from sklearn.linear_model import LogisticRegression
X = np.array([
(1, 1, 1, 0),
(1, 0, 1, 0),
(1, 1, 0, 0),
(0, 1, 1, 0),
(0, 0, 1, 1),
(0, 0, 0, 1),
(0, 1, 0, 1),
(1, 0, 0, 1)])
m = LogisticRegression()
m.fit(X, Y)
code::
Out3: array(-0.47815958, -0.47815958, -0.47815958, 1.16980067) Out5: array(0.54564474, 0.45435526) coef_ shows -0.48 (weakly suggesting group 1) with respect to the a, b, and c axes, and +1.17 (strongly suggesting group 2) with respect to the d axis.
Logistic regression adds them together, so the decision on the axes a, b, and c wins and makes the decision that the group is group 1 by a narrow margin.
code:python
from sklearn.tree import DecisionTreeClassifier
m = DecisionTreeClassifier()
m.fit(X, Y)
code::
The reason is simple: the learning process looks for "which of the axes is the most neatly divisible" and naturally leads to the conclusion that "d is the best way to make a decision". In the identification phase, no axis other than d is looked at.
2025-06-17
This story seems to have something to do with Plurality. ---
This page is auto-translated from /nishio/東洋人はロジスティック回帰で西洋人は決定木 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.