A New Approach to Clustering Essay

1217 Words Oct 1st, 2015 5 Pages
Objective of the research
The objective of this research is to increase the attention of stakeholders on the general information concerning methods used in data reduction. The research approaches this topic from a perspective where it creates a new method of clustering data. The basis of this approach is that similar and different forms of data are grouped in their own distinct sets. The level of similarity or difference is based on certain qualities in regard to the data collected. This could be in terms of distance or weight of the items in the data. The aim of this research is create new forms of clustering data that corrects some of the gaps that are created by the old existing methods data clustering. This research aims at ensuring
…show more content…
Method of developing data reduction methods
One of the methods that is used in generating data reduction method is through considering a set of data which is X and a probability measure in data X is P. In this case, we assume that the method is C which refers to representation space, where g: X C, where for there are two functions for each set K
Tk:X × X R,
Vk:C× C R,
Then the entire (x,y) € X×X,
Tk(x, y) = Vk(g(x), g(y)).
In the above derivations, the intention is to cluster data X. Assuming that data is grouped on the basis of function P and C. the g is the function that connects each element towards a particular name. (TK, Vk) represents the constants in the analysis. Therefore, {g(x): x € X} is the reduction X. However, in some cases one could consider functions that seem to be favorable to the research. In this case,
I: C R, I (g) =min
In situations where constraints are preventive and it stops the existence of functions. There is a relaxation process. Assuming that K= {1,…, N, the relaxation process will involve
J c K,
Tk(x,y) = Vk{g(x), g(g)} K€ J,
I (g) + λ J (g) = min; WK=0 if k € J
This section investigates some of the data reduction techniques that have been derived because of the above methods of forming data reduction formulas.

T= {1 if ∂ (x, y) ≥ or = U
{0 otherwise
V=0 if g(x) = g(y); 1 otherwise,
The indication

Related Documents