Skip to content

Latest commit

 

History

History
11 lines (10 loc) · 899 Bytes

README.md

File metadata and controls

11 lines (10 loc) · 899 Bytes

Classifying Textual Emotions using Neural Networks

As the amount of textual data grows at an exponential rate, it becomes increasingly vital to construct models to assess text data automatically, prompting a slew of text categorization studies. The Neural Network (CNN) has recently been used for text classification and has demonstrated to be fairly successful.

Parameters:

 Loss : poission,  Optimizer : Adam,  Activation functions : Sigmoid,  Metrics: msle  Output Layer: 6.

Outcome:

The validation process consisted of several phases that probed the parameter training methods described above. In this section, we specify the results of each phase. In order to assess the quality of the prediction, we used several figures of merit. In this study, the performance keras model tested good. With the default configuration, loss (1.0176) whereas, obtained accuracy (0.5482).