Abstract:
This undergraduate dissertation describes the project which implements a methodology for forecasting crude oil price. This project is a Web-based Application, the user will communicate with the system through web pages using web browser. In this 3-tier architecture Project there is a server-side component which will be responsible for database and synchronization services. The system is based on three main processes as historical data collection, forecasting and results (Graphs). The scope of the project encompasses both server and client-side functionalities.
The web application maintains a data warehouse for storing all the data required by the forecasting tool for the prediction of oil prices. The data is downloaded in the form of Excel files from online sources and is loaded into the Staging Database. Once the data is loaded into the Staging Database, cleansing is performed on the data i.e. data is normalized and rearranged. After the cleansing of data, the data is loaded into the Data-vault Database. This is a non-volatile (permanent) data storage. From here the data is fetched for Neural Network training, Forecasting, Reporting and generating Graphs.
JavaScript and AJAX are used on client side as scripting language. On server side, Windows Server 2012 is used as operating system and database is managed using Microsoft SQL server 2012. Asp.net framework 4.5 is used as server side scripting language.
