A Level of Useful Inference upon a Given Idea - Essay Example

Section/# The fact of the matter is that data miners by their very definition can utilize data sets to draw a level of inference upon otherwise hidden patterns. By seeking to extrapolate the way in which information can provide a level of useful inference upon a given idea, topic, or study, the d understanding of different data mining techniques is an issue of primal importance. This is true for a wide variety of different disciplines to include that of the medical profession. As such, analysis and data mining within such a field can help to emphasize and/or underscore certain otherwise hidden consistencies. As a function of this particular type of inference, the researcher and subsequent users can draw a powerful level of understanding and create the possibility for medical resolutions and improvements that would otherwise not be a possibility. With respect to this particular project, datasets from five hospitals within the nation of the United Arab Emirates will be analyzed. The information which was ultimately collected, tabulated, and analyzed related to the records of diabetic patients. Accordingly, the next step of the research was to evaluate a host of different data mining methods which could be of use with regards to understanding the overall pattern of the given datasets. As a final step, the project is intended to extract a level of useful inference with regards to the medical knowledge to include rules and /or patterns and seek to include them in an overall research that could ultimately be published as a means of showing how data mining techniques have the potential to lead to knowledge discovery and generation. As a function of performing these experiments, WEKA was utilized. As a function of using WEKA to draw a level of inference upon the medical data, it was able to determine and populate a lot of missing values by treating these values as separate attributes and encoding them appropriately. The end result of such an analysis has revealed a high number of hidden patterns that could not be elaborated upon with the use of normal analysis. As a function of all of this level of analysis, the end consequence can and should be to have a high level of improvement upon the potential for medical resolutions to both extant as well as unknown issues. Acknowledgement This particular project has only been able to be completed due to the tireless help and assistance of a wide variety of different individuals. I would like to take this particular opportunity to thank each and every one of them for the differing levels of assistance they have provided. 1.1 Supervisor Motivation When I had first begun to consider the long list of projects that were available, I had at first thought that as a software engineering specialist, it would be best to work on seeking to develop a java interface or a particular web-browsing application. However, once I considered this to a further degree, I began to realize that there was a more definitive need within the current and existing knowledge base that is represented within the Medical Data Mining arena. Due to the fact that I had long considered the techniques that could help to integrate a greater level of understanding of hidden patterns within databases, I opted for this route. Dr. M. Saraee was extraordinarily useful and helpful in seeking to motivate and direct y project as well as research into this particular field. As a function of his encouragement, this project has been able not only to improve upon my research skills but also provide a valuable and measurable impact on the prospects of future work that I might be able to engage upon. 1.1.2 Personal Motivation Over the past several years, while studying for a degree in Software Engineering, I was introduced to a few of the concept of document analysis and pattern recognition. As a result of this, one of the main motivations that I developed was to accrue a deep and complete understanding of data mining and pattern recognition. One of the course that I was able to integrate with during this time concerned the management, development and attributes of advanced database systems. It was in this particular course that I was introduced to the term “Data mining” for the first time. As a function of this introduction and preliminary study into what this represented, I came to to a full appreciation for how this field of study could have a profound and measured affect on the way that automated systems behaved. As such, this particular project has given me the ability to have the opportunity research data mining, and seek to find more efficient and useful means whereby this information can be utilized to predict and underscore, identify and reduce the prevalence of disease within the society. It is note that within my very own country of origin there is a severe shortage of both computers and individuals within a high level of programming experience. This naturally means that those individuals that have a knowledge and familiarity with pattern recognition, databases, and data mining are also quite low as compared to other universities, and professional settings around the world. As a function of this low level of expertise, this has given me yet a further motivation to pursues this particular course of analysis in seeking to provide a unit of analysis which could not only be beneficial to the medical community as a whole but to my own nation as well. Analysis of Requirement Within this brief section I will seek to identify and discuss some of the requirements that this particular Project will feature Utilizing data mining packages and means to draw a level of inference upon otherwise hidden patters that exist within medical data. Ensuring that all classification methods display all extant information, rates of error, and calculations that may have been utilized for creating the given dataset Ensuring that the basic principles of classification have explicitly been followed Utilizing histograms as a means of clearly explicating the behavior of a wide array of attributes which will be presented within each dataset Utilizing decision tress and graphical representations of methods and results to enhance the ways in which evaluation can occur Ensuring that the investigation is clear for the reader by discussing the techniques in a slow and gradual manner while at the same time ensuring that a concise and understandable approach is engendered. Read More