datacamp-logo
Session Ready
Exercise

DNN classifier using Keras API

The Banknote_Authentication dataset contains information based on genuine and forged banknotes. Extracted image features from banknotes include kurtosis, wavelength, skewness, and entropy. The Class attribute signifies whether a banknote was real (0) or forged (1).

You'll use this data to create a DNN Classifier using the Keras API. In this activity, you'll create a sequential model and train, test, and evaluate the model using the Bank Note Authentication dataset to explore how accurately you can detect a forged banknote.

For this exercise, the training and testing datasets have been split for you (banknote_authentication_train and banknote_authentication_test).

Instructions 1/3
XP
  • 1
  • 2
  • 3
  • Usekeras_model_sequential and name it ourdnnmodel.
  • Add three Dense Layers, where the first layer's shape is equal to the number of columns in train_x, the second uses relu activation, and the final layer has 1 unit.