Please wait a minute...
, No 2

■ Cover page(PDF 3234 KB) ■  Table of Content, February 2023(PDF 33 KB)

  • A New Method of Identifying the Prognostic Factors of Hepatocellular Carcinoma Patients
    Liwei Chen, Jianhao An, Mingxin Du, and Kai Su
    2023, 19(2): 85-93.  doi:10.23940/ijpe.23.02.p1.8593
    Abstract    PDF (268KB)   
    References | Related Articles
    Hepatocellular carcinoma (HCC) is a common malignant tumor. The prognostic of patients after operation involves many factors contributing to survival, and the long-term survival is unsatisfactory. Importance measures can be used to deal with other virtual systems to identify the critical factors influencing a result. In this paper, importance measures are used to analyze the prognostic factors influencing the survival time of HCC patients. Survival time and influencing factors are treated as systems and components to build a prognostic Bayesian network (BN) model based on a real-world data set. The importance values of factors are calculated using a BN-based importance calculation method. The sequence of factors based on importance measures is compared with the sequence obtained by Cox regression model to prove the effectiveness of importance measures in dealing with the prognostic influencing factors analysis of HCC patients.
    Impact of Real Time Fraud Prevention on Online Resale Platform using Machine Learning and Device Fingerprint Techniques
    Bhagirath, Neetu Mittal, and Sushil Kumar
    2023, 19(2): 94-104.  doi:10.23940/ijpe.23.02.p2.94104
    Abstract    PDF (712KB)   
    References | Related Articles
    The continuous rise in online resale transactions is associated with increased fraudulent activities in modern world economies. Customers are financially and emotionally impacted due to these fraudulent activities. To detect and prevent online fraudulent activities, there is a need for an efficient and impactful fraud detection system. Real-time fraud prevention on an online resale platform is a crucial aspect of ensuring a secure and trustworthy platform for both buyers and sellers. This leads to an increase in user trust, customer satisfaction, and reliable brand equity. In this paper, a real-time fraud detection technique is proposed with a live data stream, device fingerprint (DFP) algorithm, and machine learning techniques to improve the latency of fraudster detection and minimize the impact of fraudster activities. Real-time fraud detection mainly works on repeat fraudsters with multiple user accounts. Fraudster behavior and pattern analysis is conducted to build multiple statistics rules, seed fraudster detectors, machine learning models, and DFP mappers. The improved proposed model is quantitatively verified by the performance measuring parameter. From the result analysis, it is concluded that DFP-based real-time fraud detection increases the efficiency of detecting fraudsters and decreases the latency as compared to other conventional methods. It significantly prevents the customers from fraudulent activities and enhances the overall customer experience to a certain level.
    Quality Enhancement of Recommendation using Improved Triangle Ratings
    Devendra Gautam, Anurag Dixit, Latha Banda, Harish Kumar, Purushottam Sharma, and Chaman Verma
    2023, 19(2): 105-114.  doi:10.23940/ijpe.23.02.p3.105114
    Abstract    PDF (641KB)   
    References | Related Articles
    Recommender Systems are a potent technology used in many social networking sites. Personalized recommender systems are an added method for improving the quality of recommendation and customer’s requirements. There are many kinds of techniques available to get personalised recommendations such as Content based, Collaborative filtering and Hybrid filtering. In these mentioned techniques, the most popular CF technique is used to enhance the accuracy of RS with some shortcomings such as sparsity, scalability and cold start user problems. To enhance the quality of collaborative filtering using tagging, the proposed approach IUGT-Jaccard-ITR used may target the issue of cold start user or item problems in recommendation.
    SongRec: A Facial Expression Recognition System for Song Recommendation using CNN
    Shalaka Prasad Deore
    2023, 19(2): 115-121.  doi:10.23940/ijpe.23.02.p4.115121
    Abstract    PDF (243KB)   
    References | Related Articles
    The music has special connection with emotion of the person. One's mood can be improved by it in a special way. The classification of the emotion of music is a difficult research area because human perception is subjective. The emotional response of the user is closely related to the music recommendation system because most music is selected based on the listener's mood. Many studies have been conducted to determine how to identify emotions using various methods. These techniques have been useful in evoking the subject's feeling using a variety of devices and other hardware that can be quite expensive and inaccurate. On the other hand, observing the person’s facial expression can be quite helpful in accurately identifying their mood or feeling. Hence the main goal of the proposed system is to identify an individual's facial emotions effectively in order to make appropriate music recommendations. The proposed system makes use of Convolutional Neural Networks (CNN) to train facial dataset to recognize various emotional reactions. This trained model is used to detect mod of the person based on facial expressions and recommend song related to that emotion. The proposed system is also optimizing the results using fuzzy classification. The results demonstrate the effectiveness of the proposed methodology.
    Image-Based Crop Disease Detection using Machine Learning Approaches: A Survey
    Shikha Choudhary and Bhawna Saxena
    2023, 19(2): 122-132.  doi:10.23940/ijpe.23.02.p5.122132
    Abstract    PDF (340KB)   
    References | Related Articles
    Agriculture is a significant contributor in the global economy, as it is the key source of income for about 50% of the global population. Statistics reveal that in the financial year 2020-2021, agriculture held a share of 20.2% in the Indian economy. Destruction of crops due to diseases can cause major losses to agricultural production, thereby adversely impacting the economy. Additionally, crop diseases are a major concern for farmers as well, as they suffer a drop-in production which badly impacts their livelihood. Thus, disease detection in crops plays a significant role with regards to production of quality agricultural products. In addition to being time-consuming, the traditional methods of crop disease detection involve a large amount of expertise. To fasten up the disease detection in crops, there is a need to automate the disease detection process. Farmers are looking for instant, real-time-based, and non-destructive disease detection methods. Automated disease detection utilizing machine learning approaches can help reduce huge crop losses every year. Machine learning can be a great tool for disease identification at an initial stage, thereby enabling the conduction of preventive actions on time. This paper represents an extensive review of approaches for crop disease detection using machine learning with the aim to explore, assess, and classify the recent developments in crop disease detection.
    Software Fault Prediction using K-Mean-Based Machine Learning Approach
    Ashima Arya and Sanjay Kumar Malik
    2023, 19(2): 133-143.  doi:10.23940/ijpe.23.02.p6.133143
    Abstract    PDF (463KB)   
    References | Related Articles
    Software fault prediction is one of the most essential measures employed to evaluate software quality. There are many testing processes that are utilized to predict software faults, and among them, Black Box Testing (BBT) is used to predict faults without knowing the internal functioning of the application. In this article, a K-mean-based Machine Learning (ML) approach is explored to predict faults in software projects. The proposed model is divided into four phases. In the initial phase, the attributes of OOPs metrics that contribute to the accurate prediction of software faults are identified. In second phase, similarity between the metrics attributes is analysed using Cosine, Jaccard, and hybrid similarity measures. In the third stage, clustering of the correlated metrics attributes is performed using K-means as a clustering approach. At the last stage, Neural Network (NN) is applied as the ML approach for training and later on used for validation of the designed model. The comparative analysis is performed against the BBT and the existing work in terms of Positive Rate (TPR), Positive Predictive Value (PPV), F-score, and accuracy. The designed software fault prediction model using ML approach shows an overall classification accuracy of 94.3%.
    Hybrid Outlier Detection Strategy and Weighted Decision Matrix Ordinal Classifier for CKD Severity Prediction
    P. Antony Seba and J. V. Bibal Benifa
    2023, 19(2): 144-154.  doi:10.23940/ijpe.23.02.p7.144154
    Abstract    PDF (523KB)   
    References | Related Articles
    This article investigates the chronic kidney disease (CKD) dataset to observe meaningful insights through statistical data analysis. It is aimed to introduce a hybrid outlier detection method and a weighted decision matrix (WDM) ordinal classifier for CKD severity prediction. Attention is focused to discover the insights and to draw conclusions for feature extraction, data pre-processing, and feature selection by considering the domain knowledge, data distribution, and relationship among the variables. A hybrid approach is proposed with skewness of each variable. Interquartile range and standard deviation are introduced to handle the outliers, which are detected using univariate analysis. The various ways the values are missing in the training dataset are considered for imputation. Statistical and supervised learning approaches are utilized for selection of optimal features. The proposed outlier detection method identifies 1% of the data instances, which are extreme far points. The proposed WDM ordinal classifier model for predicting the severity of CKD is robust to feature selection, which yields an accuracy of 94.61% using the optimal features given by RFE.
ISSN 0973-1318