Question

我需要有关Keras神经网络的帮助

我没有编程方面的知识。我想让神经网络对数据进行分类（很少有目标分类变量和大约88个预测变量）。我是通过修正错误和不断Google搜索来做到这一点的：

import tensorflow as tf

from tensorflow import keras

from tensorflow.keras import layers

import keras.models

from keras import backend as K

from keras.layers import Activation, Dense

import matplotlib.pyplot as plt

import numpy as np

import pandas as pd

df = pd.read_excel('C:/Users/quad/Desktop/obradjeni podaci/Obradjeni podaci za TF.xlsx')

df.head(2720)

data = pd.read_excel('C:/Users/quad/Desktop/obradjeni podaci/Obradjeni podaci za TF.xlsx')

train_data = pd.read_excel('C:/Users/quad/Desktop/obradjeni podaci/Obradjeni podaci za TF.xlsx')

train_df = pd.read_excel( 'C:/Users/quad/Desktop/obradjeni podaci/Obradjeni podaci za TF.xlsx' )

test_data = pd.read_excel('C:/Users/quad/Desktop/obradjeni podaci/Obradjeni podaci za TF.xlsx')

train_X = train_df.drop(columns=['Dalikonzumiratecigare'])

from keras.utils import to_categorical

train_y = to_categorical(train_df.Dalikonzumiratecigare)

from keras.models import Sequential

model = Sequential()

n_cols = train_X.shape[1]

model.add(Dense(100, activation='relu', input_shape=(n_cols,)))

model.add(Dense(100, activation='relu'))

model.add(Dense(100, activation='relu'))

model.add(Dense(8, activation='softmax'))

model.compile(optimizer=tf.train.AdamOptimizer(0.001),

loss='categorical_crossentropy',

metrics=['accuracy'])

train_labels = train_data.pop('Dalikonzumiratecigare')

test_labels = test_data.pop('Dalikonzumiratecigare')

train_data = data.sample(frac=0.8,random_state=0)

test_data = data.drop(train_data.index)


train_X = train_X.transpose(92)

test_X = test_X.transpose(92)

看来我无法解决这个问题：输出：ValueError：transpose（）的pandas实现不支持'axes'参数

请帮助我。我很绝望：（

Answer 1

train_X和train_Y是熊猫DataFrame对象，而不是numpy ndarray对象。在ndarray对象上调用 transpose方法。您是在DataFrame对象上调用它的，因此会出现错误。要从熊猫ndarray中取出DataFrame，请使用：

X = train_X.values
Y = train_Y.values

Jupyter Notebook，Keras，神经网络的Tensorflow中的ValueError

1 个答案: