A-data-mining-approach-on-the-analysis-of-road-safety-in-Great-Britain

RTA data collected are huge, multi-dimensional and heterogeneous. Moreover, the data may be incomplete and contain erroneous values, which makes the data analysis a daunting task. The target data for this study was collected by the Department for Transport, GB. Several data mining techniques such as handling an imbalanced dataset, factor reduction and prediction algorithms such as Naïve Bayes, Decision Tree, Random Forest, Logistic Regression, Support Vector Machines (SVM) were carried out to perform an effective data analysis that could potentially support the transport department in devising better precautional measures to minimize the road accident occurrences in Great Britain. Moreover, the idea of chaining two different algorithms was attempted by identifying the significant attributes through Random Forest technique and feeding them as input to other ML algorithms. In addition, the key factors that influence these road collisions were identified and presented.

Building-a-data-warehouse-for-improving-the-sales-on-Northwind-database-using-SQL-Server

The idea was to manage the inventory efficiently based on the sales history by identifying the products that were in surplus amounts because of poor sales and targeting on their sales; A star schema approach was devised to build the data warehouse using SQL Server Management Studio to create dimension tables and fact table; Tableau and Power BI were used to visualize the data; The Northwind model was implemented in Neo4j, a graph database to compare how it fares against the traditional ER modelling techniques. Further graph information retrieval and analysis were demonstrated using cypher query language for Neo4j.