VAERS-Covid19-MachineLearning

Project idea from Kaggle: COVID-19 World Vaccine Adverse Reactions by Ayush Garg. Project for class CS584ML

Note:

Raw data and merged dataset is not uploaded due to large size. Original data can be found: https://www.kaggle.com/ayushggarg/covid19-vaccine-adverse-reactions?select=2021VAERSVAX.csv

License

CC0: Public Domain

script.py

The script include five steps from data processing to graphing to training model and to using model for prediction.

1 Preprocess data

• needs data files inside rawdata directory
• merge 3 datasets into 1
• preprocess text data into numbers, number to text mapping are saved under category directory

2 Visualize raw data

• needs data files inside rawdata directory
• graph data into scatter plots

3 Train model

• needs merged dataset file
• split merged dataset into feature and target datasets
• Train model with 4 choices:

1. Stochastic Gradient Descent Classification
2. Logistic Regression Classification
3. K Neighbors Classification
4. Neural Network Classification with MLP After training, model are saved under models directory.

4 Using model to predict patient outcomt

• needs trained and saved files inside models direcotory and input file
• print prediction category and predicted outcome for each patients in a list of 0 and 1.