Accepted Papers
- Data Mining Through Neural Networks Using Recurrent Network
Gaurab Tewary,Myself, India
ABSTRACT
With the development of database, the data volume stored in database increases rapidly and in the large amounts of data much important information is hidden. If the
information can be extracted from the database they will create a lot of profit for the organization, & the technology of mining information from the massive database is known as data mining. The main purpose of data mining is to gain insight of the data, and extract knowledge (meaningful patterns) from the data. Data mining tools can forecast the future trends ,detect fraud information. There are many technologies available to data mining practitioners, including Artificial Neural Networks, Genetics, Fuzzy logic and Decision Trees.
Neural network is a parallel processing network which generated on the basis of the research on biological neural network, according to the features of biological neurons and neural network . Its advantages such as high affordability to the noise data, low error rate and high accuracy. The Recurrent Structure is also known as Auto associative or Feedback Network ,in Feedback Network, the signal travel in both the directions by introducing loops in the network.It consists of Competitive networks & Hebbian's rule etc. Many data mining practitioners don't use Neural Networks due to their black box nature, even though they have proven themselves in many situations. To remove the black box nature of neural networks we shall use ESRNN (Extraction of Symbolic Rules from Artificial Neural Networks), it consists of Weight Freezing, Pruning, Rule Extraction etc. In this paper, we shall discuss , Why the artificial neural networks should be used as a preferred tool by data mining practitioners ,How we can use ESRNN to overcome the "Black Box Nature" of Neural Networks, What is the role of Recurrent Network in neural
- Comparision Of Decision Tree Algorithms In Predicting Student's Absenteeism In An Academic Year
G. Suresh1 , K. Arunmozhi Arasan2, S. Muthukumaran3 ,1St. Joseph's College of Arts and Science, 2Siga College of Management and 3Siga College of Management and Computer Science , India
ABSTRACT
Educational data mining is used to study the data available in the educational field and bring out the hidden knowledge from it. Classification methods like decision trees, rule mining can be applied on the educational data for predicting the students behavior. This paper focus on finding the best algorithm which gives the best result to find out the reason for leave taken by the student in an academic year. The first step of the study is to gather student's data by using a questionnaire. We collect data from 123 students who were under graduate from a private college which is situated in a semi-rural area. The second step is to clean the data which is appropriate for mining purpose and choose the relevant attributes the classification is done using the gender attribute. In the third and final step, we used three different Decision tree induction algorithms namely, ID3 (Iterative Dichotomiser), C4.5 and CART (Classification and Regression Tree) for the same data set and the results were compared. We used the comparison results to decide the best classification algorithm in predicting the reason for absenteeism of students for the classes.
- Mood Rcognition using Indian music : A Survey
Vishwa D. Joshi, Ravindra A. Vyas, Dharmsinh Desai University, Nadiad, India
ABSTRACT
The study of mood recognition in the field of music has gained a lot of momentum in the recent years with machine learning and data mining techniques and many audio features contributing considerably to analyze and identify the relation of mood plus music. In this paper we consider the same idea forward and come up with making an effort to build a system for automatic recognition of mood underlying the audio song's clips by mining their audio features and have evaluated several data classification algorithms in order to learn, train and test the model describing the moods of these audio songs and developed an open source framework. Before classification, Preprocessing and Feature Extraction phase is necessary for removing noise and gathering features respectively
- HPPDFIM-HD: Transaction Distortion and Connected Perturbation Approach for Hierarchical Privacy Preserving Distributed Frequent Itemset Mining over horizontally-partitioned Dataset
Fuad Ali Mohammed Al-Yarimi Alandalus University for Science & Technology , Sana'a- Yemen
ABSTRACT
Many algorithms have been proposed to provide privacy preserving in data mining. These protocols are based on two main approaches named as: the perturbation approach and the Cryptographic approach. The first one is based on perturbation of the valuable information while the second one uses cryptographic techniques. The perturbation approach is much more efficient with reduced accuracy while the cryptographic approach can provide solutions with perfect accuracy. However, the cryptographic approach is a much slower method and requires considerable computation and communication overhead. In this paper, a new scalable protocol is proposed which combines the advantages of the perturbation and distortion along with cryptographic approach to perform privacy preserving in distributed frequent itemset mining on horizontally distributed data. Both the privacy and performance characteristics of the proposed protocol are studied empirically.
- A K-Means Clustering Algorithm For Predicting Student's Absentiesm In An Academic Year For Categorical Datasets
G. Suresh, R.Karthikeyan, S. Muthukumaran, St.Joseph's college of Arts and Science, India
ABSTRACT
Clustering algorithms are a useful tool to explore data structures and have been employed in many disciplines. Educational data mining is used to study the data available in the educational field and bring out the hidden knowledge from it. Currently many educational institutions are facing problems with the lack of attendance among the students. The aim of this paper is to predict students absenteeism in an academic year using k-means algorithm for categorical data. The first step of the study is to gather student's data by using a questionnaire. We collect data from 123 students who were under graduate from a private college which is situated in a semi-rural area. The second step is to clean the data which is appropriate for mining purpose. K-means algorithm uses Maximum Component Analysis (MCA) to convert the data for categorical to numeric and Equilidian distance is used to measure the distance and clustering is done with this distance. This knowledge is used to identify the reason fo r the leave taken by the student and help to improve the quality of the environment and also to improve the performance of the student.
- Enhancing Automated Diagnosis Of Ecg Using Hl7 Medical Device Communication
Monica Maiti, Computer Science and Engineering, Saveetha Engineering College, Chennai,Tamilnadu.
ABSTRACT
Electrocardiogram data are analyzed and stored in different formats, devices and platforms; hence it is tedious to display those data unless the user has access to that particular software of each particular device. In the current scenario, diagnosis of electrocardiogram data includes all the waves and signals, nevertheless due to the presence of noise in the Electrocardiogram data, the results of the diagnosis might be misleading. Therefore, in order to avoid these issues, a system is proposed which performs an image validation of histogram check to rectify the noise acquired in the input Electrocardiogram image as well as tune up and improve the quality of the original image. Furthermore, the use of Health Level 7 as the medical data exchange standard integrates electrocardiogram waveforms as well as data descriptions of the disease. Consequently, an automated XML report will be generated, representing specific cardiac abnormalities through the development of ontology.
- Towards Interesting Rare Itemset Mining using Tree Structure
Urvi Bhatt, Pratik Patel, Computer Science and Engineering, Gujarat Technological University, Gujarat, India
ABSTRACT
Pattern mining methods describe valuable and advantageous items from a large amount of records stored in corporate datasets and repositories. While mining, literature has been almost singularly focused on frequent itemset but in many applications rare ones are of higher interest. Example of such application can be a medical dataset where rare amalgamation of prodrome plays a vital role for the physicians. As rare items contain worthwhile information, researchers are making efforts to examine effective methodologies to extract the same. In this paper, we make an effort to analyze the complete set of rare items for finding the most off all rare association rules from the dataset. Proposed approach uses Maximum constraint model for extracting rare items. A new approach is efficient to mine rare association rules which can be defined as the rules containing the rare items. Based on the study of relevant data structures of the mining space, our approach utilizes the tree structure to ascertain the rare items. Finally we have tried to demonstrate that, this approach is more virtuous and robust than the existing algorithms.
- Report Visualization Of Inter-Related Data Warehouse Using Integration Algorithm
Sylvia Irish.S, Computer Science and Engineering, Saveetha Engineering College, India.
ABSTRACT
Data mined from the single source is analyzed and a report is generated using first order logic in the existing system. But this system fails to support inter-related storage of data. Hence, an Interactive analysis of data has been proposed, by allowing data to be summarized and viewed in different ways. Data that can be modeled as dimension attributes and measured attributes. Measure attributes include measure of some value that can be aggregated upon .e.g. the attribute number of sales relation. Mediator based data integration algorithm has been implemented for generalization & suppression of data's in the data warehouse. Acute evasion technique to handle bulk data has been deployed further in our proposed approach. Hierarchy on dimension attributes of interrelated data has been proposed where it lets dimensions to be viewed at different levels of detail. The proposed methodology is also useful at the beginning of the visualization process.
- Sentiment Analysis
R. Indhu, Ezhilarasi. L, Anna University Chennai, India
ABSTRACT
Mining affective information from text and provides author's feeling towards the particular entity is very important because there is no physical contact to find out the meaning of the text in context. A number of methods for classifying text into positive and negative but semantic approach help to find out the sentiment of the text based on context. This paper describes lexicon-based semantic approach to classify the movie review into positive and negative depending on the polarity scores. The task is performed in three steps: (1) extracting subjective sentences; (2) creating dynamic-sentiment base; (3) classify the review as positive when polarity score greater than 0 otherwise negative.
Courtesy
Technically
Sponsored by
