Machine-Learning-Approach-to-Crash-and-Crime-Analysis-using-R

The overall goal of this research is to compare and evaluate the performance of machine learning algorithms. The research applied five machine learning algorithms to analyse Town of Cary’s crash and crime data. Using R Studio (v. 1.1.456), R Libraries and Microsoft Office Excel (v.2016), data cleaning and other manipulations were carried out on the Town of Cary, North Carolina’s crash and crime data. All data were sourced from the US Open Data Portal.

A-MapReduce-Approach-to-Weather-and-Exchange-Rate-Analysis-

Data manipulation like attribute renaming, record filtering, missing record update and so on, were carried out on weather and exchange rate data sourced from Marine Institute and the Central Bank of Ireland respectively using Anaconda Navigator (v. 5.3, Jupyter Notebook, Python 3.6), R Studio Packages (v. 1.1.456) and Microsoft Office Excel (v.2016). Using Java Programming Language, map and reduce functions were defined; the two datasets were mapped, Linear and Multiple Linear Regression analysis were carried out to establish the relationship between weather, some of its parameters and euro exchange rates for six countries. All data were analyzed in parallel, using installed Hadoop on a Linux Ubuntu 16.4 instance running on OpenStack. From the analysis, it was discovered that wind direction and sea temperature have a positive correlation with Euro exchange rates for Australian Dollar, Great Britain Pounds, Hong Kong Dollars, Japanese Yen and US Dollars while Air Pressure has a negative correlation with the same. Also, air temperature has a negative correlation with Euro exchange rates for Great Britain Pounds and US Dollars but has a positive correlation with the Euro exchange rates for Hong Kong Dollars, Australian Dollars and Japanese Yen.