[FIXED] Preparing data for sequential model (machine learning)

Issue

I take data from a file

y = np.genfromtxt('dataY.txt', dtype=np.float32) #input data for target dataset
x = np.genfromtxt('dataX.txt', dtype=np.float32)#input data for input dataset

I split the data accordingly:

xtrain, xtest, ytrain, ytest = train_test_split (x,y,test_size=.2)
train_data = tf.data.Dataset.from_tensor_slices((xtrain, ytrain))
valid_data = tf.data.Dataset.from_tensor_slices((xtest, ytest))

The dataY.txt file is made by 1000 rows. Each row includes 30 numbers that I want the NN to guess after the training, given the input X.

The dataX.txt file is made by 1000 rows, one for each Y. Each row includes 100 numbers.

Question: how do I make the following code work?

model = Sequential()
#whad do I need to write in the following line(s)?
model.add(Conv2D(100,(7,7)))
model.add(LeakyReLU())
model.compile( loss='mse', metrics=['mse'])
model.fit(train_data, epochs=10, validation_data=valid_data)
ypred = model.predict(x)

ERROR:

ValueError: Input 0 of layer sequential is incompatible with the layer: : expected min_ndim=4, found ndim=2. Full shape received: (100, 1)

Solution

It looks like you're using a convolutional layer for 1D data. Try dense layers instead:

model = tf.keras.models.Sequential()
model.add(tf.keras.Input(shape=(100,)))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(30))
[...]

Also I'd recommend printing the shapes of x, y, xtest, xtrain, etc. to ensure you're feeding in the correct dimensions.

(Source: https://keras.io/api/layers/core_layers/dense/)

Answered By - Skrt