Please wait a minute...
, No 3

■ Cover page(PDF 3.07 MB)■  Table of Content, March 2021  (PDF 92.4 KB) 

  • Original article
    Material Modeling of Epoxy Granite Composite by Analytical Model and Regression Analysis
    S. Nallusamy
    2021, 17(3): 253-262.  doi:10.23940/ijpe.21.03.p1.253262
    Abstract    HTML   PDF (603KB)   
    References | Related Articles

    This research work mainly focuses on the material modeling of epoxy granite composite with the help of experimental methods. Generalized mixture rule (GMR) for particulate composite was proposed by screening to estimate a specific mechanical property of epoxy granite with regards to their properties, volume fractions, and microstructures (PVFM) of component. In GMR relation, the effects of microstructures were expressed by means of a scaling fractal parameter j. Taguchi’s design of experiments was applied to plan the number of experiments. The investigations were carried out based on flexural and damping test of epoxy granite specimens with dimension 125x12x6mm and young’s modulus with specimens of dimension 50x50x50mm. Analytical values of specific mechanical property and varying volume fraction were presented for j values 0.25 to 0.1. The experimental results obtained from different tests were plotted over the analytical graph, which further helped to fix the value of j for each effective property. From the results, a unique j value of 0.3 was finalized for epoxy granite with the j fixed for each effective mechanical property. Regression analysis was applied to establish the empirical relation between effective and material properties for experimental results. Comparison between the analytical model from GMR and regression model from experimental results was carried out to validate the mathematical model.

    Arrhythmia Classification Algorithm based on SMOTE and Feature Selection
    Tianhao Wang​, Peng Chen​, Tianjiazhi Bao​, Jiaheng Li​, and Xiaosheng Yu
    2021, 17(3): 263-275.  doi:10.23940/ijpe.21.03.p2.263275
    Abstract    HTML   PDF (677KB)   
    Figures and Tables | References | Related Articles

    An arrhythmia is commonly deemed as a life-threatening disease. It is better to detect symptoms of arrhythmia earlier, as this can be more beneficial for relevant treatment. Presently, classification research on arrhythmias by machine learning is mainly dependent on data extracted from ECGs. However, some defects can still be found in arrhythmia data, such as class imbalance, strong correlation among features and high dimensions. All these defects have the potential to incur classification inaccuracy. In an attempt to solve the above problems, an arrhythmia classification algorithm is proposed here based on SMOTE and feature selection. Firstly, dataset oversampling was performed by SMOTE to erase class imbalance; then, K-part Lasso was utilized to select the existing redundant features; finally, recursive feature elimination (RFE) and random forest (RF) are combined together to form a feature selection method, RF-RFE, for the purpose of selecting optimal features. In this way, feature sub-sets were acquired and further adopted to carry out classified evaluations and comparisons of four classification algorithms. It has been demonstrated by UCI arrhythmia datasets-based experiments that 89 of the 279 features in the raw data are selected by the proposed arrhythmia classification algorithm. Such selected features that serve as the optimal feature sub-set are used for classification. Moreover, the accuracy of the RF classification reaches 98.68%.

    Code Confusion in White Box Crowdsourced Software Testing
    Run Luo​, Song Huang​, Hao Chen​, and MingYu Chen
    2021, 17(3): 276-288.  doi:10.23940/ijpe.21.03.p3.276288
    Abstract    PDF (632KB)   
    Figures and Tables | References | Related Articles

    In recent years, crowdsourcing software testing as a new testing service mode has been widely concerned. However, white box crowdsourcing software testing is often regarded as an insecure testing service mode. The main threat comes from unknown attacks in the crowdsourcing environment, which leads to the risk of source code leakage in white box testing. This paper discusses the weakness of white box software testing in crowdsourcing software testing, as well as the possible mode of attack. This paper proposes to use code obfuscation technology as a solution to this kind of attack and analyzes the impact of code obfuscation technology on crowdsourcing testing. This paper is the first attempt at using code obfuscation technology in white box crowdsourcing software test task protection.

    Exponential Moving Average Modelled Particle Swarm Optimization Algorithm for Efficient Disassembly Sequence Planning towards Practical Feasibility
    Anil Kumar Gulivindala, M.V.A. Raju Bahubalendruni, S.S.V. Prasad Varupala, and Chandrasekar Ravi
    2021, 17(3): 289-298.  doi:10.23940/ijpe.21.03.p4.289298
    Abstract    HTML   PDF (546KB)   
    References | Related Articles

    The application of artificial intelligent (AI) algorithms in disassembly sequence planning (DSP) has attracted a lot of research attention recently due to their effectiveness at solving combinatorial problems. Particle swarm optimization (PSO) is the most widely preferred AI algorithm for obtaining an optimal solution for the DSP problem. However, the solutions generated from traditional PSO have limitations due to its converging nature at local optima. In this research, an attempt has been made to improve the workability of PSO by integrating it with the exponential moving average (EMA) method. The optimality function is designed to reduce disassembly effort by considering tool changes, gripper changes and directional changes as parameters. A case study has been performed by testing the proposed EMA-PSO method on the 11-part industrial product. Obtained results are revealed that the diversity control is greatly achieved by the operators employed in the disassembly attributes. The effectiveness of the proposed EMA-PSO method is confirmed by making a comparative assessment with traditional PSO and other existent AI methods at different population sizes.

    LSTM and RNN to Predict COVID Cases: Lethality’s and Tests in GCC Nations and India
    Razia Sulthana A.​, Arokiaraj Jovith​, and Jaithunbi A. K.
    2021, 17(3): 299-306.  doi:10.23940/ijpe.21.03.p5.299306
    Abstract    PDF (410KB)   
    Figures and Tables | References | Related Articles

    The spread of COVID across world countries is better handled by applying learning algorithms. Machine learning and deep learning algorithms can be applied to analyze the effects of COVID in multidimensional ways. This paper brings a detailed study of the COVID cases, deaths and tests across five of the GCC countries and India. The proposed method analyzes the COVID count against the population density of each of the countries. An analysis of the raw count would only give a false impression, whereas a population-based comparison gives the exact measure of the effect of COVID. As India is a densely populated country, the number of precautionary steps taken by the country against the population count needs to be measured for accurate prediction. Recurrent Neural Network and Long Short-term memory are used to predict the future cases, deaths and tests of India. A time span of 20 days is used in the prediction. In the sense that ith day to (i+20)th day values are taken to predict the (i+21)thday values. The accuracy of the LSTM model designed with multiple hidden layers is notable and the prediction error is minimal.

    Polynomial Curve Fitting-based Early Room Reflection Analysis using B-Format Room Impulse Response Measurements for Ambient Sound Reproduction
    Vuppala Swathi and Sandeep Chitreddy
    2021, 17(3): 307-313.  doi:10.23940/ijpe.21.03.p6.307313
    Abstract    PDF (469KB)   
    Figures and Tables | References | Related Articles

    A polynomial curve fitting based early room reflection parameter analysis is discussed in this work. Particularly, the spatial directions of ground reflections that are usually harder to extract from Omni-RIR are analyzed using measurements of B-Format microphones. A simulated model is initially presented with the ray approximation of sound propagation. Subsequently, a recently developed parameterization approach called Reverberant Spatial Audio Object (RSAO) is discussed for early reflection parameter extraction from measured B-Format RIRs. Parameters obtained from Simulated and RSAO methods are analyzed for 10 different source receiver distances. A publicly available classroom B-Format RIR database is used for extracting RSAO parameters for 10 distances. Polynomial curve fitting is performed on the parameters obtained from both methods. The optimal order that can generalize the 10 samples parameters is also obtained as part of this work. The results show that the method is suitable for obtaining parameters for non-measured directions from a sparse set of measurements of B-Format RIRs.

    Original article
    Hand Sign Recognition using CNN
    D. Bhavana, K. Kishore Kumar, Medasani Bipin Chandra​, P.V. Sai Krishna Bhargav​, D. Joy Sanjanaa, and G. Mohan Gopi
    2021, 17(3): 314-321.  doi:10.23940/ijpe.21.03.p7.314321
    Abstract    PDF (604KB)   
    Figures and Tables | References | Related Articles

    Our aim is to produce a model that can recognize hand gestures and signs. We will train a model for the purpose of sign language conversion, a simple gesture recognizing model; this will help people converse with people who are innately deaf and mentally disabled. This project can be implemented in several ways such as KNN, Logistic Regression, Naïve Bayes Classification, Support vector machine and can be implemented with CNN. The method we have chosen is CNN as it gives better accuracy compared to the rest of the methods. A computer program is developed using python language which is used to train the model based on the CNN algorithm. The program will be able to recognize hand gestures by comparing the input with preexisting dataset formed using the American sign Language. We will be able to convert Sign Language into text as output for users to recognize the signs presented by the sign language speaker. This model is implemented in Jypter Lab, an extension to the platform Anaconda documentation. To further improve, we will also add / integrate the inputs into black and white and take input from camera after using the method of Background subtraction. With the mask set to detect the human skin, this model will not require a plain background to function and can be implemented using a basic camera and a computing device.

    Detecting Pulmonary Embolism using Deep Neural Networks
    J Akilandeswaria, G. Jothib, A Naveenkumara, R. S. Sabeenianc, P. Iyyanara, and M. E Paramasivamc
    2021, 17(3): 322-332.  doi:10.23940/ijpe.21.03.p8.322332
    Abstract    PDF (1785KB)   
    References | Related Articles

    Medical image processing is a method to create visual representations of the internal parts of the human body such as organs or tissues which helps diagnose and monitor diseases. Pulmonary Embolism (PE) is a medical issue where there is an artery obstruction in the lungs. PE is the third most common cause of cardiovascular death and is related to multiple inherited and acquired risk factors. The earlier diagnosis of PE detection helps to increase the patient's survival. With the advancement of Artificial Intelligence (AI), deep learning has become the leading technique, as it established significant capabilities in medical image processing tasks. In this research, a popular deep learning technique called Convolution Neural Network (CNN) is used to detect the pulmonary embolism in lung CT scan images. Four different types of predefined CNN architectures such as the Inception, VGG-16, ResNet50, and Mobilenet are used to compare the performance of the CNN model. In this experiment, RSNA STR Pulmonary Embolism Chest CT scan image dataset is analyzed. The empirical results show that the Inception-based CNN model provides better results when compared to other CNN architectures.

ISSN 0973-1318