Udacity deep learning nanodegree project 5
- Software and Service used
- Project Inspiration
- File Descriptions
- Data Insights
- Licensing, Authors, and Acknowledgements
Python version 3.0. Amazon Web Service to train, evaluate and deploy the web app.
Get a sense of how to deploy a machine learning model using AWS Sagemaker, including upload train, validation and test data to S3; develope and train, validate and test RNN, then do the following as this picture indicate to deploy my web app that can judge a movie review is positive or negative.
Deploy_a_sentiment_analysis.ipynb : Jupyter notebook containing all the codes and results
report.html : notebbok in html format
train.py : model file
predict.py : predict fucntion file
index.html : web app
After deploy my web app, I generated few reviews and test it, the following are the results.
The intersting one is the last two, I said 'this movie is long and boring, but I like ...' and 'the only thing I like about this movie is it has the actress I like', the web app thinks both are NEGATIVE review and they do. Even tho I use plain and positive words like 'like', 'actress'. It feels like this web app understands my feeling, which feels funny in a way. Anyways, I only typed few reviews and there are room for improving, we could use LSTM RNN, etc....
Data: IMDb data
Authors: To the Notebook