线性回归问题 model

Question

I have this DataFrame I created, using data from basketball reference and I get the mean for each characteristic. DataFrame Data No matter which column I used to train my Linear Model, my R2 score is near 0 and the predictions are awful.

import pandas as pd
import numpy as np
from sklearn import linear_model
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import r2_score
import math


percent=math.floor(len(df)*0.80)
X=df['Mean MP'].to_numpy()
Y=df['BPM'].to_numpy()
Y=Y.reshape(Y.shape[0], )
X=StandardScaler().fit_transform(X)

X_train=X[:percent]
Y_train=Y[:percent]
X_test=X[percent:]
Y_test=Y[percent:]

model=linear_model.LinearRegression()
model.fit(X_train,Y_train)

a=r2_score(Y_test,model.predict(X_test))
a=-0.07

I think its how I create or use my DataFrame but I don't know how to correct the problem. I wish you could help me, thank you.

Answer 1

Maybe this is not the right problem for machine learning. Are you sure that there is some relation between one or more of the statistics present in your dataframe and the BPM score. Perhaps try using a multiclass classification algorithm like decision tree, using all the features, after changing BPM to categorical scores, like between 0 and -2 is a bench player, between 4 and 6 is an all star-consideration. I know nothing about basketball, and I used this link to understand things.