CSC591 Machine Learning for User Adaptive Systems Assigned Project 1: Missing Data Handling
Group ID: ALL_8 .
Group Members:
Chengyuan Liu (cliu32@ncsu.edu)
Rachit Shah (rshah25@ncsu.edu)
Sourabh Sandanshi (ssandan@ncsu.edu)
Find the python script ML_AP1_ALL_8.py to train or test in the zip file as well.
How to Run:
python ML_AP1_ALL_8.py --train #to train
python ML_AP1_ALL_8.py --test #to test
python ML_AP1_ALL_8.py --tune #to hyper parameter tune
Make sure folder contains folders train_data/train_with_missing, train_data/train_groundtruth, test_data/test_with_missing, joined_missing.csv and joined_groundtruth.csv
The zip also contains python notebook ML_AP1_ALL_8.ipynb which contains same code to train, test, and tune but also contains graph plots for exploratory data analysis.
Best way to run the notebook is using google colab. Open the notebook and colab at https://colab.research.google.com. You can find link to the colab notebook when you open the notebook locally. Or you can run locally on jupyter.
Packages required -
seaborn, fancy impute, pandas, numpy
Follow instructions in Notebook to run the code. You might need to upload the zip files to google drive. The provided joined csv files will reduce the time taken to run and combine the csv files. The file upload code paragraph needs to be run two/three times if there is any error.