ICP_Deep Learning 3 - PallaviArikatla/Python GitHub Wiki

INTRODUCTION: To work on Embedding layer in keras and to Perform sentimental analysis on the datasets given.

IMPLEMENTATION:

Question 1: In the code provided,there are three mistake which stop the code to get run successfully; find those mistakes and explain why they need to be corrected to be able to get the code run.

  • The three identified mistakes and changes to be done are
  1. The input dimension must be equal to the number of columns in the dataset given, hence we have to give value for the input_dim variable.

  1. Change the neurons in the output layer as 3 as we have three values in the label i.e., neg, pos and unsup.
  2. Make softmax activation, as it suits best for multi class classification as there are multiple divisions in the target column.

  • Calculate Loss and Accuracy.
  • Plot Loss and Accuracy.

Question 2: Add embedding layer to the model and check whether there is any improvement.

  • Prepare the data for embedding.
  • Encode the data followed by splitting the data to train and test.
  • Check for the imports and import necessary embedding layers from keras layers.

  • Add the embedding layer.

  • And plot the accuracy.

Question 3: Apply the code on 20_newsgroup data set we worked in the previous classes.

  • Follow the same procedure as above and change the data to 20newsgroup data and then change the output neurons to 20.
  • Execute the program 20newsgroup on and plot the accuracy.

Adding embedding layers accuracy got decreased.

Bonus questions:

Question 1: Plot the loss and accuracy using history object.

  • Consider loss and accuracy values from the the data given and history object and plot graph loss against accuracy.

Question 2: Predicting the fourth sample.

  • Consider fourth value in the data and draw its predicted value.