Abstract Classification is the most solved and the most used machine learning problem. In a last few decades many classification algorithms have been developed. Because of that, when classification is needed in some problem solving, the best algorithm should always be selected. The problem that is analysed in this paper is choosing classification algorithms that can be used in vertical retrieval... Classification is the most solved and the most used machine learning problem. In a last few decades many classification algorithms have been developed. Because of that, when classification is needed in some problem solving, the best algorithm should always be selected. The problem that is analysed in this paper is choosing classification algorithms that can be used in vertical retrieval system for both document and query classification. We compared SVM, Multinomial Naïve Bayes algorithm, Bernoulli Naïve Bayes algorithm and Random forest. The experiments presented in the paper, show that in the long documents classification SVM and Multinomial Naïve Bayes algorithms have a similar precision (SVM is a little better), but the Multinomial Naive Bayes algorithm correctly classified 93.14% of queries, while SVM only 22.55%. Read More Read Less

Abstract In this work, we have studied the VoIP quality signal waveform analysis. We have presented models for objective prediction of the voice quality for IP networks and also monitor the quality control of VoIP networks. We studied the methods to find the quality of audio. The study has analysed the quality of VoIP connection and indicate during quality decrease. This... In this work, we have studied the VoIP quality signal waveform analysis. We have presented models for objective prediction of the voice quality for IP networks and also monitor the quality control of VoIP networks. We studied the methods to find the quality of audio. The study has analysed the quality of VoIP connection and indicate during quality decrease. This exercise provided the chance of solving the issues in VoIP network before users are affected by VoIP specific connection problems (echo, noise or breaks in the conversation). The signal waveform envelope distortion is reviewed; practical questions of its numerical implementation are discussed. We supported the study with many illustrations. Read More Read Less

Abstract To recognise and detect the ingredients used in the tea, we have deployed Artificial intelligence techniques. The techniques that we have used are the VIS/NIR spectroscopy, measurement of color and pH. Principal components analysis (PCA). We have adapted the reduced high dimensional feature space to three specific factors. We further present the experimental results in the paper. The Back propagation... To recognise and detect the ingredients used in the tea, we have deployed Artificial intelligence techniques. The techniques that we have used are the VIS/NIR spectroscopy, measurement of color and pH. Principal components analysis (PCA). We have adapted the reduced high dimensional feature space to three specific factors. We further present the experimental results in the paper. The Back propagation Artificial Neural Network (BP-ANN) and Kohonen’s Self organizing map (SOM) shown the studied samples with an accuracy of 99.4% and 98%, which ensure the use of these techniques for future also. Read More Read Less