It is often seen that business houses in clusters are more successful than those businesses doing identical trade or commercial ventures but are separated from each other, either through geographical distances or other barriers which do not offer their clustering. Not only are clusters more successful in terms of profitability and growth indicators but they are also well-positioned in terms of good management structures and high share value with respect to the companies who are not clustered.
It is often found that in many countries the situation of certain similar industries is in close geographical proximity to each other. For instance, computer firms are found in the US in Silicon Valley and Bangalore in India, and they are proliferating exponentially in the recent times. This geographical proximity also gives rise to clustering and has a positive impact on the company's growth since all the members of the cluster would be able to contribute to the economic welfare of the State to which it belongs. The concept of clustering is fundamentally to determine the essential grouping or collation of data, and in the contest of business enterprises could be used to consider important aspects like revenue generation, profit making corporation and the main location aspects of the business. Through this research it is tried to focus upon the important aspects that work in the case of Clustering and how this aspect impacts upon the economy of the country.
Clustering could be used for a variety of applications depending upon the desired results or objectives and could help in studying areas of interest like biology, insurance, seismic recordings for earthquakes, World Wide Web (www) where it is necessary to determine the pattern of behaviour in seemingly discrete and unorganized data.
For instance in determining census reports, a lot of mass of data from all over the districts are collated and gathered and for arriving at the correct distributional patterns among the vast mass of data and to provide intelligent and coherent analysis and interpretations for the data . This could be done through the use of cluster formation which would distribute the similar data in identical formatting and thus help in achieving the desired ends and conclusions regarding the distribution and behaviour patterns of the data under study. The main determinants of clustering would be therefore in terms of
Ability to make valid measurements and coherent analysis
Assign attributes to the data based on the findings of clustering
Formulating data with the various designs for implementation of decisions
Established the least requirement for domain knowledge and determining the inputs that could provide valid outputs measurements and the basis for arriving at logical conclusions
It is to be noted that the domain is not concerned with the order of the inputs recording since the processing would be based on similar identification and not other criteria
High dimensionality would have to be accorded to the data being researched since this would impinge upon the final results
It is also necessary that the data on clustering has high degree of usability in the academic context
There are different types of clustering and they could be seen in the contest of Exclusive clusters, overlapping clusters, hierarchical clusters and probable clusters.
The Euclidean distance would be the