Name of the project
Amazon category prediction
Tool used:
• Python
Packages used:
• NumPy
• Pandas
• Matplotlib
• Tensorflow
• Scikit learn
• Keras
Dataset
https://www.kaggle.com/code/parichart/sentiment-analysis/data
Data collection:
• The product data is obtained from the Amazon product data set on Kaggle, which contains information on millions of products sold on Amazon.
Data cleaning and preprocessing:
• The product data is cleaned and preprocessed to remove any missing or irrelevant data, and to extract features such as the length of the product title and the number of words in the title.
Model selection and training:
• Using recurrent neural network(RNN) loss: 0.0755 - accuracy: 0.9771 - val_loss: 0.3084 - val_accuracy: 0.9201 was obtained as final result.
• Using Long short-term memory (LSTM) loss: 0.0247 - accuracy: 0.9920 - val_loss: 0.2634 - val_accuracy: 0.9430 wsa obtained as final result.