Publication

Abstract : Television is becoming an important part of our everyday life as a mass medium for communication. With its dramatic and demonstrative strength, to change it is a true source of knowledge, education and entertainment. It is convincing to take into account the promotion of social values and norms in a civilized society. The structure and shape of the shape play a major role. Here, we propose a new method called multi-class classification with the help of big data analytics. In this study, we have provided a framework for comparing the quality of various classification methods using statistical simulation when individuals belong to one of the two groups that are mutually exclusive. Here we compare naïve Bayes classification, multi-layer perceptron classification and decision tree classification as a test case. From the results, we found that the classification accuracy of multi-layer perceptron classification is higher than the other two in analysing the television viewing behaviour. The data for the study is collected through a direct questionnaire survey. Here, accuracy is assessed over correctly and incorrectly categorized instances. The analysis found that MLP is the best algorithm to predict TV viewing behaviour.