MINING THE WORLD DATA AND PREDICTING HEALTH EXPENDITURE OF A COUNTRY

Predicting health expenditure of a country is aimed to study various indicators from the world development data website and to come up with useful data and design a predictive model that helps the government to take important measures related to health development.

We have used various indicators from world bank data sets available. The purpose of our study is to construct a model that has the information or the countries around the world for few years. The information we considered for our study include total population, urban population, crude birth rate, crude death rate, incidence of TB,prevalence of anemia etc.

In this study, we have applied various data mining tasks, performed exploratory data analysis, correlation techniques, missing data imputation, outlier handling, predictive model that classifies whether a country is going to have low, medium, high or very high health expenditure based on various predictor variables.

Team Members:

1. JYOTHIRMAYI PANDA
2. SNEHA VAISHNAVI
3. JAI KIRAN DUVVU

Results

This slideshow requires JavaScript.

Social Mining with R

In this Project i had performed Twitter Data Analysis of Narendra Modi  the 14th and current Prime Minister of India using R language.

R packages used:

1. twitteR – Querying Twitter
2. ROAuth – Authentication with Twitter
3. tm – Text mining
4. ggplot2 – Graphics Visualization
5. RColorBrewer – Color Mapping
6. wordcloud – Building a word map with most frequent words
7. sentiment – Performing a sentiment analysis for the tweets and determining the score of tweets

Results

Time Series Forecasting with R

In this Project i did time series forecasting of Euro-USD exchange rates with the with the ARIMA and STL models. The data used are historical currency exchange rates from January 1999 to June 2016 provided by the European Central Bank.

R packages used:

1. forecast

Results