Exploring-the-non-linear-relationship-between-Crimes-and-GDP-using-Generalized-Additive-Models

This repository contains the script and figures of the conference paper selected for presentation at the Latin American Conference of Computationa Intelligence 2018. The abstract of the paper is as follows: Crime is an important social and economic problem of South Africa. Though certain categories of crimes are of serious proportions, yet on aggregate the overall crime situation in the country has considerably improved in the last decade or so. A number of previous studies across other countries have shown a positive or negative relationship between crime and economic growth. On a microeconomic/provincial scale, this paper studies the relationship between various categories of crimes and economic growth using the non-linear modeling technique of Generalized Additive Models. Such a modeling approach helps in understanding how various categories of crimes complexly affect GDP.

Early-estimation-of-protest-time-spans-Using-NLP-Topic-Modeling

Protests and agitations have long used as means for showing dissident towards social, political and economic issues in civil societies. In recent years we have witnessed a large number of protests across various geographies. Not to be left behind by similar trends in the rest of the world, South Africa, in recent years have witnessed a large number of protests. This paper uses the English text description of the protests to predict their time spans/durations. The descriptions consist of multiple causes of the protests, courses of actions etc. Next we used unsupervised (topic modeling) and supervised learning (decision trees) to predict the duration of protests. The results are very promising and close to 90% of accuracy in early predicting of the duration of protests.

Facebook_analytics_on_NDTV_using_R_and_Shiny

This project is a visual and NLP based interactive analytics to understand what NDTV shares in facebeeok in terms of posts, images, and videos; and how its followers/subscribers react to such postings. It consists of two modules: PART 1: is where NDTV's facebook posts are fetched using the Rfacebook package. The script used in fetching the data is named "ndtv_getPage.R" The data fetched from script is saved as "ndtvpage.rds" Please fill in your app_id and app_secret in the fields marked XXXX. PART 2: The shiny scripts are ui.R and server.R Please set the appropriate paths for the readRDS() and the scan() functions. positive.csv and negative.csv are repository of positive and negative sentiment words.

Nature-and-characteristics-of-Indian-political-parties

This work is a visualization of the various characteristics of the political parties and its members of the 16th Lok Sabha of the Indian Parliament. The data can be downloaded from http://www.prsindia.org/mptrack/rajyasabha under the **Download Data** section. For a vast majority of the Ministers of Parliament(MPs), their offices started from 18-May-14 and most they are presently holding their offices. The author neither claims nor is responsible for the authenticity of the data.

Interactively-forecasting-Global-temperature-using-R-and-Shiny

This work is a extensive interactive visualization as well as forecasting tool to forecast global monthly temperature. Choice of country and state (for which the forecasting is required) is menu driven and based upon dynamic subsetting of data. In addition, various graphs and parameters pertaining to the model builiding, comparison of predicted and actual values and future forecast are all based upon click of the reelvant action buttons. The time series algorithm used is Exponential Smoothing, and fairly good results are obtained for a large combinations of countries and states.