SHM Publisher Journals
Not a member yet
315 research outputs found
Sort by
Naive Bayes and KNN for Airline Passenger Satisfaction Classification: Comparative Analysis
Get PDFAir transportation is vital due to technological advancements and globalization. It is affordable and accessible worldwide, providing efficient services to reach destinations globally. This discussion focuses on full-service airlines that offer online-based services. Previous research indicates that available facilities and services influence passenger satisfaction. Previous research on customer satisfaction showed a correlation between satisfaction and services without accurate figures. In the present study, the customer satisfaction figure is measured using the Naive Bayes and K-Nearest Neighbour (K-NN) algorithm to obtain a tested level of accuracy. In this analysis, we will compare the effectiveness of Naive Bayes and K-NN algorithms in classifying airline passenger satisfaction. The results show that the accuracy of the Naive Bayes method of the two algorithms is higher than the K-NN method. The accuracy value of the Naive Bayes method is 84.48%, while the accuracy value of the K-NN method is 65.38%. From the test results, the precision value for Naive Bayes is 82.25%, and K-NN is 67.35%. Furthermore, the recall value for Naive Bayes is 82.43%, and K-NN is 74.33%
Analysis of Micro Structure, Porosity Disability and Wear Resistance with Volume Variation from Riser On the Engine Cover of Electric Motors
Get PDFThis research to analyze the microstructure, porosity defects, and wear resistance of aluminum casting in the presence of variations in the volume of risers. The method used is the pre-experimental method of the One-Shot Case Study type, because in this study a treatment will be carried out and the results will be observed. The treatment that will be carried out is the addition of variations not using risers, variations of riser cylinders with a volume of 2826mm3, variations of riser cylinders with a volume of 4710mm3, variations of riser cylinders with a volume of 6594mm3. Microstructure testing used the Meji Techno IM 7200 test tool. Wear testing used the Ogoshi High Speed Universal Wear Testing Machine (Type OAT-U). The data analysis used is descriptive analysis to provide an overview of the research subject based on data from the variables obtained from the group of subjects studied. The best microstructure is shown by the cast specimens with the riser volume variation of 6594mm3 as evidenced by the formation of a more dense and even structural phase. The best porosity results were shown by the specimens that were cast with a volume variation of the riser 6594mm3 of 37.97%. The best wear resistance results in variations of the volume riser 6594mm3 with wear values of 0.51x10-7mm2/kg
Lung cancer classification using convolutional neural network and DenseNet
Get PDFLung cancer is a condition that has a major impact on public health. Convolutional Neural Network (CNN) and DenseNet approaches are suggested in this study to aid lung cancer detection and classification. In various fields of pattern recognition and medical imaging, CNN and DenseNet have demonstrated their efficacy. In this study, radiology images from individuals with lung cancer were used to create a set of medical lung images. The findings show that lung cancer can be accurately classified into malignant and benign from radiological images using CNN and DenseNet architectures, with a parameter accuracy of 99.48%. This research contributes to the creation of a deep learning-based system for detecting and classifying lung cancer. The findings can be the basis for creating a more accurate and productive lung cancer diagnostic system
An expert system on diagnosis of mental diseases
Get PDFMental disorder is one of the most serious problems in today's time. Mental disorders can be classified into different sub-disorders according to changes in human behavior and mental condition. According to reports one out of seven people suffered from mental disorders. In this research paper, our main emphasis is to build an expert system that diagnoses people based on their symptoms, so people can diagnose themselves early before going to the doctor. Expert Systems are one of the most important applications in artificial intelligence that solves complex problems without human help. We provide different rules, facts, and relationships among different symptoms in our knowledge base, from which users can query their problems and get their results. We used SWI-prolog to build an expert system. There are a few types of disorders, such as mental disorders, neurodevelopmental disorders, eating disorders, etc
Analysis of quality of service (QoS) wi-fi network in UNNES digital center building using wireshark
No full textThe need for the internet is a very absolute target in today's all-digital era. The traffic of information that is so dense and always dynamic every second makes everyone want speed in capturing information circulating. The speed in gathering information in this all-digital era cannot be separated from the internet and networks. UNNES Digital Center is one of the facilities owned by Semarang State University which is used as a digital-based learning center to support the realization of the Smart Digital Campus. The availability of qualified network services at the UNNES Digital Center is needed to support the all-digital-based student learning process. This research was done to find out how fast and good the quality of the internet network provided by the UNNES Digital Center is. In the research conducted, the network analysis step uses the Quality of Service (QoS) method. In obtaining research data that will be used as a basis for analyzing throughput, packet loss, delay, and network jitter, Wireshark software is used as a tool. The research results show that the quality of the Digital Center's internet network is very good and very adequate for digital learning activities. This is evidenced by a network throughput value of 6122.37 /kbits/s, a packet loss value of 0.7%, a delay of 214 ms with a moderate or quite good value and jitter = 0.511 ms
Securing audio chat with cryptool-based twofish algorithm
Get PDFTechnology and the internet are growing very rapidly in society making it easy for people to share information and communicate with each other. However, the security of such data or information is something that should be highlighted. The utilization of technology and the internet has many security gaps that can make data or information vulnerable to being stolen and even misused. Valuable data or information is very likely to be accessed by unauthorized people. On the other hand, data and information are prone to be illegally altered and even duplicated. In connection with various possible data or information security issues, it is necessary to do a data security. The purpose of this study is to secure audio chat using a Cryptool-based Twofish algorithm. Based on research conducted, the security process with encryption and decryption simulations was successfully carried out on audio chat. Audio chat sent via IP Address can be encrypted into ciphertext and can be decrypted back into audio at a speed of 15.78 kB/s and the resulting size is also still the same, which is 160 packets. This Twofish algorithm proved to be well usable because the size and quality of chat audio generated from decryption is still the same as audio chat before it is encrypted
Security improvement of aes algorithm using s-box modification based on strict avalanche criterion on image encryption
Get PDFCommunication is something that cannot be separated from humans as social creatures. Images are the most commonly used visual communication in today's era. On the other hand, sending images via wireless networks is very vulnerable to piracy. AES, as one of the best cryptographic algorithms, can be applied as a solution. Even so, the AES algorithm still has weaknesses, which are weak against linear attacks and differential cryptanalysis. One solution to overcome the weaknesses of the AES algorithm is to use a stronger S-box. One of the methods to measure the strength of an S-box is the Strict Avalanche Criterion (SAC). The dataset is divided into four categories based on the image type and size of the pixels. Data that has been encrypted using the proposed algorithm will be compared with data that has been encrypted using the standard AES algorithm. Cipherimages (encrypted data) are tested using histogram analysis, information entropy, and sensitivity analysis. The results obtained from cipher image testing are differences in histogram analysis testing in grayscale and color images. The information entropy value is 0.000131583% better than the AES standard, the NPCR is 0.17613% better than the AES standard, and the UACI value. 0.211148% better than AES standard in sensitivity analysis testing. Based on these data, the proposed algorithm has a higher level of security than the standard AES algorithm on image encryption
Analysis of public opinion sentiment against COVID-19 in Indonesia on twitter using the k-nearest neighbor algorithm and decision tree
Get PDFCOVID-19 has become an ongoing disease pandemic across the globe. The need for information makes social media such as twitter a place to exchange information. This tweet can be used to see public sentiment towards COVID-19 in Indonesia. Sentiment analysis classifies opinions from tweets that have been processed and classified into different sentiments, namely negative, neutral, or positive. The aim of this paper is to find the algorithm that has the best accuracy. The researcher proposes to compare the K-Nearest Neighbors (KNN) and decision tree algorithms to be used in the classification of sentiment data from tweets related to COVID-19 that took place in Indonesia. The results of the evaluation of performance metrics concluded that the decision tree algorithm has a higher level of accuracy than KNN. Decision tree produces accuracy = 0.765, error = 0.235, recall = 0.76, and precision = 0.767 which is better when compared to KNN which produces accuracy = 0.69, error = 0.31, recall = 0.66, and precision = 0.702
Performance evaluation and measurement of SMEs king of honey using the green SCOR method
Get PDFAssessment of the application of the concept of Green Supply Chain Management (GSCM) in each business unit is something that is important to do considering that currently environmental issues are a very urgent matter. Therefore, this study will focus on assessing the GSCM process and measuring performance to determine the value of GSCM performance in King of Honey SMEs. The method used in this research is Green SCOR with 6 management processes, namely plan, source, make, delivery, return, and waste. From the research conducted, the priority level of GSCM indicators and the value of GSCM performance on King of Honey SMEs are generated. The results of this study showed that the total performance value of GSCM King of Honey in September was 86.03, October was 86.45 and November was 86.48
Business process modeling at steak restaurant using business process model and notation
Get PDFThe complexity of business processes occurring today makes the company try to find ways to describe its business processes. Business processes are not only an operational standard but also become one of the determining factors for the smooth use of time and costs in a business unit to be more efficient. With good business processes, it makes the flow of information faster so that it can help in making the best decisions in the organization. The business process modeling that will be explained further in this study is the order and procurement business process at steak restaurant using the Business Process Model Notation (BPMN) approach. This research was conducted using a qualitative descriptive method with the aim of observing the business unit to help analyze and make improvements to its business processes. Several series of processes were carried out, namely business identification and modeling with bizagi modeler and process reengineering to produce recommended new business process models that could be beneficial for business units, namely the recommended automation in the form of the use of mobile applications, remote, and database systems to support the effectiveness of the order to cash and procure to pay processes