GitHub - courtneyjean/NaiveBayesCRPovertySurvey_MATLAB: Using Naive Bayes to predict income levels from the Costa Rica Household Poverty Survey

courtneyjean / NaiveBayesCRPovertySurvey_MATLAB Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Using Naive Bayes to predict income levels from the Costa Rica Household Poverty Survey

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
NB and Ensemble Methods for predicitng income for the Costa Rican Household Survey.pdf		NB and Ensemble Methods for predicitng income for the Costa Rican Household Survey.pdf
README.txt		README.txt
cleandata.m		cleandata.m
grid_search_nbv5.m		grid_search_nbv5.m
macrof1.m		macrof1.m
train.csv		train.csv

Repository files navigation

How to run code for the Using Naïve Bayes to predict income levels from the Costa Rica Household Poverty Survey

Authors: Courtney Irwin (Matthew Stewart contributed to this project using Random Forrest as a comparator method.  The Naive Bayes code here is my own.)

MATLAB Version:MATLAB2018b

Required Files are:
 - cleandata.m : a function to clean the data
 - grid_search_nbv5.m : the main file for running the naive bayes algorthim
 - macrof1.m: a function to calculate the macrof1 from a confusion matrix
 - train.csv: the full dataset from the model; 
		downloaded from https://www.kaggle.com/c/costa-rican-household-poverty-prediction/data

To run the code and reproduce the results, run (only) the following files in MATLAB2018b:
- grid_search_nbv5.m : the main file for running the naive bayes algorthim

The other function files will be called from within these files.