CONCLUSIONS

Chapter XIV - Financial Benchmarking Using Self-Organizing Maps Studying the International Pulp and Paper Industry
Data Mining: Opportunities and Challenges
by John Wang (ed)
Idea Group Publishing 2003

	Brought to you by Team-Fly

In this study, financial information for 77 companies in the international pulp and paper industry for the years 1995 2000 was collected using the Internet as a source of information, and a financial database was created. A number of financial ratios were selected and calculated based on the information in the database. Then, a data-mining tool, the self-organizing map, was used to perform a financial competitor benchmarking of these companies.

This tool offers a number of benefits over other alternatives. Using self-organizing maps, we can compare much larger amounts of data than by using spreadsheet programs. It is also useful for exploratory data analysis, where patterns are not known a priori. Unlike traditional clustering methods, like k-means clustering, the number or size of clusters does not have to be decided a priori, and visualization is much more graphic. A drawback with self-organizing maps is that the person who runs the analysis has to be an experienced user of neural networks. Also, like all neural networks, self-organizing maps are "black boxes," i.e., the user does not see how the network works or what it does, only the result. Finally, technical issues, such as standardizing the data, are very important for the final result.

The results of the study provide further evidence that the self-organizing map is a feasible and effective tool for financial benchmarking. The results are easy to visualize and interpret, and provide a very practical way to compare the financial performance of different companies. The discovered patterns were confirmed with existing domain knowledge.

	Brought to you by Team-Fly