Data Mining and Knowledge Engineering

Ayesha Butalia ¹, Maya Ingale ², Swapnil Patil , Manjiri Bangali , Bhushan Dusane

Affiliations
1 Maharashtra Institute of Technology, College of Engineering, Pune, IN
2 School of Computer Science and Information Technology, D.A. University, IN

Abstract

Gesture recognition can be seen as a way for computers to begin to understand human body language, thus building a richer bridge between machine and humans than primitive text, user interfaces or even GUI’s. Gesture recognition method used in this paper aims on recognizing hand postures as well as facial expressions. Here, facial expressions are used as context for the gesture recognized that helps in giving additional information about the gestures performed. Thus the information obtained from this static posture defines a particular gesture.
We also have described another method which uses previous frame as its context and recognize the gesture performed. Many a times, a gesture recognized can have multiple meaning, or its actual meaning cannot be easily recognized by just extracting features from one frame. So, in such situations we can use previous frame of the gesture performed and extract its features. The information obtained from these features can then be used to avoid the ambiguity in meaning and the gesture can be recognized accurately. These gestures then can be used to make some complex decisions, ultimately driving an application.

Keywords

Grey Scale Manipulation, Smoothing, Edge and Feature Detection, Face and Gesture Recognition.

Full Text

     

Ayesha Butalia ¹

Affiliations
1 Computer Department, MIT College of Engg., Pune, IN

Abstract

The issues of Real World are Very large data sets, Mixed types of data (continuous valued, symbolic data), Uncertainty (noisy data), Incompleteness (missing, incomplete data), Data change, Use of background knowledge etc. Lot of knowledge related to the application can be generated through these large data sets. Rough set is the methodology which can be used to deduce rules from these data sets. The main goal of the rough set analysis is induction of approximations of concepts. Rough sets constitute a sound basis for KDD. It offers mathematical tools to discover patterns hidden in data and hence used in the field of data mining. Rough Sets does not require any preliminary information as Fuzzy sets require membership values or probability is required in statistics. Hence this is its specialty. Two novel algorithms to find optimal Reducts of condition attributes based on the relative atttribute dependency, out of which the first algorithms gives simple Reduct whereas the second one gives the Reduct with minimum attributes. This project highlights on the case study of mushroom which consists of twenty two attributes depending on which the decision is taken whether the mushroom plant is edible or poisonous. The technique of Reduct is very useful as when tested, through the algorithms, the twenty one attributes, excluding the decision attribute gets reduced to two to three attributes.

Keywords

Data Mining, Rough Sets, Reducts, Mushroom.

Full Text