Data mining (DM) for stands English Data Mining is the process of extracting useful knowledge and Understandably, previously unknown from large quantities data stored in different formats. The tools Data Mining predicts future trends and behaviors, allowing the business the making decisions.
There are terms frequently used as synonyms for data mining. One of them called “analysis (Smart) data “, which tends to place more emphasis on techniques statistical analysis. Another term widely used, and more related to data mining, is the removal or “Knowledge Discovery in Databases or KDD, acronym in English”.
Although some authors use the terms Data Mining and KDD interchangeably, as synonyms, there are clear differences between the two. Thus most authors agree refer to the KDD as a process consisting of a set of phases, one of which is data mining. Accordingly, the data mining process is only in the implementation of a algorithm to extract data patterns and the process is called KDD includes full pre-processing, mining and post-processing of data.
The KDD according to is the extraction Automated knowledge or interesting patterns, not trivial, implicit, previously unknown potentially useful predictors of information from large databases.
Obtaining patterns and rules in the process Academician of the University of the Science computer using techniques of mining of data.
Since the implementation of a group of techniques Data Mining as clustering, the trees decision and algorithms of learning inductive, is to sort students according to academic performance and later find hidden patterns and rules that characterize, based on the relationships established between the center of origin students, educational level of parents and province home with their academic performance in first grade in college. These results may improve the process of and raise academic quality of the education at the University of the Sciences Computer (ICU).
This intends to classify the Investigation students of the University of Informatics Sciences According to their academic behavior using a set of Data Mining Techniques like clustering, decision trees and inductive learning algorithms. The main goal is to work find hidden patterns and rules that define this behavior, based on the relationship established between the scholarship level of the student’s parents and their origins with their academic grades in the first year of their career. These results help to can improve the quality of the academic process in the ICU.
The University of Informatics Sciences (UCI) account from the 2006-2007 school year with an enrollment of about 10 000 students from all provinces and municipalities, with the most various social and academic backgrounds, without, far, studies have been conducted to evaluate the influence of these factors in its further formation. Because these factors are not taken into account when make the process of attracting students new entry to college, or to give as the necessary follow-registered, which may lead in extreme conditions to Withdraw from the center. While in other cases are left to identify students with greater potential, which could form part of projects or groups of research or just arm the with faculty information convenient for them to provide attention differentiated students in sake of promoting full development of their capabilities and thus giving effect to goal University’s primary, which is to train professionals of computing better and better prepared.
All the information personal and teaching students from five years ago is digitized and held in that historic provide greater utility the of traditional reports.
This is why the University is necessary count on methods efficient and automatic to explore the great bases data processed quickly and reliably information to find patterns knowledge appropriate to solve a problem.
This is why the main objective of this work seek to determine the link between the origin and social background of students in the ICU with their academic performance by application of clustering techniques and Association Rules Mining Data.