This is a project that performs Sentiment Analysis on a Dataset of 25000 movie reviews. We're trying to classify the reviews into positive (1) and negative (0) sentiments.
You can find more information on kaggle.
The challenge was to complete the analysis withing 1 week. (06.04.20-13.04.20)
The work is done in a Jupyter Notebook, where the final model can be created and saved.
We mainly used a LogisticRegression model, that achieved close to 90% accuracy on our test set. Other models are also included and a general Outlook for improving the test result is given.
Tobias Becher
- GitHub TB-DevAcc
- LinkedIn Tobias Becher
- Hackerrank TB_DevAcc
CoderSchool for their help, support & mentorship
Titlephoto by Pietro Jeng from Pexels
Titlephoto in the Notebook from NinjaEssays
Copyright © 2020 Tobias Becher.
This project is MIT licensed.