colleagues, actually I am kind'a new to NN, but hard trying..

I have data:

Index: 40073 entries (excluded from training, UID)

Columns: 484 entries dtypes: bool(468), float64(2), int64(13), object(1)

I used only 478 arguments. The Y is moneySpend which can be >= 0

The code is below:

newDropped = df.drop(["moneySpend","userAgent", "secondsToBuy", "hoursToBuy", "daysToBuy", "platform"], axis = 1)

x_train, x_test, y_train, y_test = train_test_split(newDropped, df["moneySpend"], test_size=0.25, random_state=547)

model = Sequential()
dnn1.add(Dense(16, input_dim=478, activation='relu'))
dnn1.add(Dense(8, activation='relu'))
dnn1.add(Dense(1, activation='linear'))

model.compile(loss='mse', optimizer='adam', metrics=['accuracy'])
tb_callback = tf.keras.callbacks.TensorBoard('./logs', update_freq=1)
history = model.fit(x_train.astype('int'),  y_train.astype('int'), validation_split=0.33, epochs=100, batch_size=5, callbacks=[tb_callback])

I have kind'a big losses, and no idea what number of neurons to use at each Dense layer (should I use Dense or LSTM?) and what proper number of layers and activation functions to use.

Regularization is making training worse, so I turned it off.

Any good advices for me, if you can?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.