Data Mining and Knowledge Engineering

S. Charles ¹, D. Charles Raj ¹

Affiliations
1 Department of Computer Science, St. Joseph College, Trichy, IN

Abstract

Big data is an emerging technology, which enables to determine the facts from various repositories using analytics. In an earlier stage, the researchers are not able to handle huge amount of data. The various techniques are employed for finding the accuracy of huge amount of data. In this context, a machine learning technique helps to predict the information from big data repositories. In machine learning techniques, there are supervised, semi-supervised and unsupervised learning schemes are used to predict the information. In this paper, supervised techniques and their issues are considered for the discussion in various applications of big data.

Keywords

Big Data, Supervised Classification, Decision Tree, Support Vector Machine.

Full Text

     

S. Charles ¹, Y. Sunil Raj ¹, R. Bastin Jesu Raj ², S. Antony Joseph Raj ¹

Affiliations
1 Department of Computer Science, St. Joseph's College (Autonomous), Tiruchirapalli-620002, IN
2 Millennium Software Productions, Chennai-600001, IN

Abstract

Classification is the most frequent technique for mining data in various domains. The accuracy measures are used to predict the accuracy, sensitivity, specificity and ROC values. This research focuses on finding the accuracy of machine learning techniques using medical data sets, which adopts the supervised learning techniques for analyzing the medical datasets. The study determines the classification accuracy of four medical datasets such as hepatitis dataset, heart disease dataset, and breast cancer data set and diabetes data set. The classifiers are employed to find out the correctly classified instances, incorrectly classified instances and Error rate. This Investigation helps to determine the accuracy of the classification analysis and it leads to find the disease affected stages. Also assists the medical practitioners in terms of treating the patients based on the stages of disease.

Keywords

Accuracy, Classification, Decision Making, Error Rate.

Full Text

     

L. Arockiam ¹, S. Charles ¹, M. Amala Jayanthi ²

Affiliations
1 Department of Computer Science, St. Joseph's College, Tiruchirappalli, IN
2 Department of Computer Applications, Kumaraguru College of Technology, Coimbatore, IN

Abstract

The education's objective focuses on students. The objective of education is fulfilled once the students are moulded in their knowledge, skills and attitudes under efficient supervision of educators. According to the Bloom's theory on learning activity, the learning activity of a person not only improves one's knowledge and mental skills (i.e. Cognitive) but also his emotional areas (i.e. Affective). A person's is defined as the amalgamation of emotional, attitudinal, behavioural responses (i.e. Affective). Happiness is one of the, s (i.e. Affective). To be a well being and good personality is more important, and education helps a individual to grow as good personality and a wellbeing. Educational Data Mining is one of the Blooming applications in educational sector. It helps to understand better the learning activity of students and their overall involvement in the activity, focused on the further improvement of the quality and the productivity of the educational system. This research is to study the influence of the Personality traits and Emotional Happiness on the academic performance of the students according to Bloom's theory. Eysenck Personality Inventory and Criterion Reference Model used to determine the personality of the students. Oxford Happiness Inventory and Criterion Reference Model is used to evaluate the student's emotional happiness. This paper researches the impact of Personality traits and Emotional Happiness in students' learning process based on the supervised and unsupervised techniques to analyse students' dataset. Multi-Layer Perceptron and EM clustering Technique is employed. To cluster the students based on the Personality, Emotional Happiness Level and Performance, Multi-Layer Perceptron and EM clustering Technique is employed .The study determines the association between students' Personality, Emotional Happiness and Performance through descriptive and predictive modelling using mapping or function. It shows that there exists a positive correlation between student's Personality, Emotional Happiness and Performance. This research allows the educators to understand students' Behavioural, Attitudinal, Emotional Growth during the learning activity as a Personality and a wellbeing and provides appropriate training for improving their proficiency in academics.

Keywords

Multi-Layer Perceptron, Expectation Maximization (EM) Clustering, Criterion Reference Model, Bloom's Taxonomy, Affective Domain, Eysenck Personality Questionnaire, Personality Types.

Full Text

     

S. Charles ¹, L. Arockiam ¹

Affiliations
1 Department of Computer Science, St. Joseph's College (Autonomous), Tiruchirappalli, Tamil Nadu, IN

Abstract

Feature selection is one of the process data mining tasks. This process finds the optimal feature subset using machine learning and evaluation criteria. Several techniques are used to find the optimum features in artificial and text databases. In this paper, the machine learning based methods are classified as unsupervised learning, semi-supervised learning and supervised learning, which find the various features in the text databases, web databases and gene databases. The evaluation criteria based methods are categorized as Filter, Wrapper and Hybrid approach, which are employed to discover the optimal feature set in artificial datasets. These approaches are very useful in data mining process for improving the prediction performance, reducing the cost and understanding of the features. These issues are addressed by various techniques using measures like dependent and independent criterion. This survey explores the various feature selection processes and their uniqueness for finding the optimal feature subset in term of accuracy, robustness and efficiency.

Keywords

Feature Subset Generation, Feature Subset Evaluation, Stopping Criteria, Feature Subset Validation, Dependent Criterion, Independent Criterion, Unsupervised Feature Selection, Semi-Supervised Feature Selection, Supervised Feature Selection, Filter Approach, Wrapper Approach and Hybrid Approach.

Full Text

     

L. Arockiam ¹, S. Charles ¹, M. Amala Jayanthi ¹, S. Mercy ¹

Affiliations
1 St. Joseph's College, Trichy, IN

Abstract

Students are the heart of the Educational Environment. The Objective of the education focuses on students. The Motto of the education is to refine the students’ knowledge, skills and attitudes under efficient supervision in the educational environment. As per Bloom’s theory on learning activity, a person's learning activity not only improves his/her knowledge and mental skills (i.e. Cognitive) but also his/her emotional areas (i.e. Affective). Educational Data Mining is one of the blooming applications in educational sector. It helps in better understanding of students’ learning process and their overall involvement in the process, focused on the improvement of the quality and the profitability of the educational system. Personality is defined as the combination of emotional, Attitudinal, Behavioural Responses of an Individual. Personality is an intrinsic factor of affectivity. Emotional Happiness is an affective category. This research is to study the influence of the Emotional Happiness of students on their personality. Emotional Happiness of the students is evaluated using Oxford Happiness inventory. The Personality of the student is determined by the Eysenck Personality Inventory and Criterion Reference Model. This paper finds out the impact of Emotional Happiness on the personality of the students in their learning process based on the supervised and unsupervised techniques for analysing the students' dataset. Multi-Layer Perceptron and EM clustering Technique is employed to classify the students based on the Emotional Happiness and Personality. The study determines the association between the students' Emotional Happiness and Personality through descriptive and predictive modelling using mapping or function. It reveals that there exists a positive correlation between student Emotional Happiness and Personality. This investigation allows the teaching community to understand students' Emotional Happiness and Behavioural, Attitudinal, Emotional Growth during their learning process as Personality and provide them appropriate counselling to develop them as good human persons in the society.

Keywords

Multi-Layer Perceptron, Expectation Maximization (EM) Clustering, Criterion Reference Model, Bloom's Taxonomy, Affective Domain, Eysenck Personality Inventory, Personality Types, Oxford Happiness Inventory, Emotions, Emotional Happiness.

Full Text

     

L. Arockiam ¹, S. Charles ¹, M. Amala Jayanthi ¹, A. Sahaya Mercy ¹

Affiliations
1 Department of Computer Science, St. Joseph's College, Tiruchirappalli, IN

Abstract

The soul of education environment is students. Through Education, students are cultivated in their knowledge, skills and attitudes under supervision. According to the Bloom's view on learning activity, a person‟s learning activity not only depends on his knowledge and mental skills (i.e. Cognitive) but also on his emotional areas (i.e. Affective). Educational Data mining is one of the Blooming applications in educational sector. It helps in better understanding of student‟s learning process and their overall involvement in the process, focused on the improvement of the quality and the profitability of the educational system. This research is to study the students' performance which is influenced by their emotions as per Bloom's view, especially student's Emotional Happiness (i.e. Affective). The Emotional Happiness and the performance of the students are evaluated by the Oxford Happiness Inventory and Criterion reference model. This paper finds out the impact of emotional happiness in students' learning, which adopts the supervised and unsupervised techniques for analysing the students' dataset. Multilayer Perceptron and EM clustering Technique is employed to classify the students' based on the Emotional happiness and performance. The study determines the association between the students' happiness level and performance through descriptive and predictive modelling using mapping or function .It shows that there exists a positive correlation between student Happiness and performance. This Investigation allows the teaching community to understand student's behaviour and provide appropriate training for their improvement of academic competence.

Keywords

Multilayer Perceptron, Expectation Maximization (EM) Clustering, Criterion Reference Model, Bloom's Taxonomy, Affective Domain, Oxford Happiness Inventory, Emotions, Emotional Happiness.

Full Text

     

L. Arockiam ¹, S. Charles ¹, C. Lalitha ¹, A. Sujanitha ¹, V. Arul Kumar ¹

Affiliations
1 Department of Computer Science, St. Joseph’s College, Tiruchirappalli-620002, IN

Abstract

Juvenile Plantar Dermatosis (JPD) known as sweaty sock syndrome. It affects the skin, which becomes scaly and red on the soles of the feet of the children and the young teenagers. A new framework is developed to provide an offline solution for JPD using descriptive and predictive modeling. The symptoms and stages of JPD are classified using clustering and decision tree technique. In clustering technique, the patients’ objects are categorized using symptoms. The decision tree method has been employed for finding the decision rule based on symptoms, which is used for recommendation. The recommendations are useful to the patients those who are affected by the JPD disease.

Keywords

K-Means, Clustering, JPD, Decision Tree.

Full Text