我在使用MLP进行机器学习时遇到了这个问题，我不确定是否很好。数据集（道琼斯指数数据集）具有16个属性，我想使用火车的第一季度数据和测试的第二季度数据（季度是另一个属性）来最大化“ percent_change_next_weeks_price”（其中一个属性）。当我打印预测（predictii）时，我得到类似[0.24083294 0.24083294 ........... 0.24083294 0.24083294 0.24083294]的信息，我不知道这是否正确。

from sklearn import neural_network
from sklearn.feature_extraction.text import CountVectorizer
import pandas as pd

s = pd.read_csv(r'C:\Users\Europa\Desktop\dow_jones_index.data',encoding = "ISO-8859-1")

#s=s.loc[:,['quarter','stock','date','open','high','low','close','volume','percent_change_price','percent_change_volume_over_last_wk','previous_weeks_volume','next_weeks_open','next_weeks_close','percent_change_next_weeks_price','days_to_next_dividend','percent_return_next_dividend']].apply(lambda x : x.str.strip('$'))

train = s[s['quarter'] < 2]
test = s[s['quarter'] ==2]
transftrain= []
for row in range(0,len(train)):
    transftrain.append(' '.join(str(x) for x in train))
countvector=CountVectorizer(ngram_range=(2,2))
traindataset=countvector.fit_transform(transftrain)

transftest= []
for row in range(0,len(test)):
    transftest.append(' '.join(str(x) for x in test))
test_dataset = countvector.transform(transftest)

regr = neural_network.MLPRegressor()

regr.fit(traindataset,train['percent_change_next_weeks_price'])

predictii= regr.predict(test_dataset)

print(predictii);