机器学习无法预测正确的结果

Question

我正在创建一个简单的 python 机器学习脚本，它将根据以下参数预测贷款是否会被批准

business experience: should be greater than 7
year of founded: should be after 2015
loan: no previous or current loan

如果满足以上条件，则只批准贷款。这个数据集可以从这个link:

下载

https://drive.google.com/file/d/1QtJ3EED7KDqJDrSHxHB6g9kc5YAfTlmF/view?usp=sharing

对于以上数据，我有以下脚本

from sklearn.linear_model import LogisticRegression
import pandas as pd
import numpy as np

data = pd.read_csv("test2.csv")
data.head()

X = data[["Business Exp", "Year of Founded", "Previous/Current Loan"]]
Y = data["OUTPUT"]

clf = LogisticRegression()
clf.fit(X, Y)

test_x2 = np.array([[9, 2017, 0]])
Y_pred = clf.predict(test_x2)
print(Y_pred)

我正在通过test_x2中的测试数据。测试数据是如果business exp是9，成立年份是2017，没有current/previous贷款，那么就是提供贷款。所以它应该预测并且结果应该是 1 但它显示 0。代码或数据集是否有任何问题。由于我是机器学习的新手并且仍在学习它，所以我创建了这个自定义数据集以供我自己理解。

请大家给点好的建议。谢谢

Answer 1

您应该在管道中使用 StandardScaler()

from sklearn.linear_model import LogisticRegression
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import make_pipeline
import pandas as pd
import numpy as np

data = pd.read_csv("test2.csv")
data.head()

X = data[["Business Exp", "Year of Founded", "Previous/Current Loan"]]
Y = data["OUTPUT"]

clf = make_pipeline(StandardScaler(), LogisticRegression())
clf.fit(X, Y)

test_x2 = np.array([[9, 2017, 0]])
Y_pred = clf.predict(test_x2)

print("prediction = ", Y_pred.item())
prediction =  1
print("score = ", clf.score(X, Y))
score =  0.95535

机器学习无法预测正确的结果

Machine learning not predicting correct results

python

machine-learning

scikit-learn

logistic-regression