Machine Learning
Finding Best Model
CHAT-GPT USAGE: Throughout this project, we made great use of CHAT-GPT. It was the tool that gave our project suggestions for the models early on in the project. It also guided us to the methods of performing TF-IDF scoring and Sentiment Analysis. We kept asking it ways to improve at every stage and it was crucial to our project.
Fitting Random Forest
After successful feature extraction, we explored various machine learning modules like Logistic Regression, Support Vector Machines and Random Forest Classifier. We randomly split our data into testing and training sets with an 80:20 split. After fitting the best features and running the various models, we found that Random Forest Classifier gave us the best accuracy of 0.9671 or 96.7%. In our hyper parameter tuning, we used the number of trees or n_estimators to be 100.
