Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control
Not a member yet
424 research outputs found
Sort by
Employee Attrition and Performance Prediction using Univariate ROC feature selection and Random Forest
Get PDFEach company applies a contract extension to assess the performance of its employees. Employees with good performance in the company are entitled to future contracts within a certain period of time. In a pandemic time, many companies have made decisions to carry out WFH (Work from Home) activities even to Termination (Attrition) of Employment. The company's performance cannot be stable if in certain fields it does not meet the criteria required by the company. Thus, due to many things to consider in contract extension, we are proposed feature selection steps such as duplicate features, correlated features and Univariate Receiver Operating Characteristics curve (ROC) to reduce features from 35 to 21 Features. Then, after we obtained the best features, we applied into Decision Trees and Random Forest. By optimizing parameter selection using parameter grid, the research concluded that Random Forest with feature selection can predict Employee Attrition and Performance by obtain accuracy 79.16%, Recall 76% and Precision 82,6%. Thus with those result, we can conclude that we can obtain better prediction using 21 features for employee attrition and performance which help the higher management in making decisions
Moving Objects Semantic Segmentation using SegNet with VGG Encoder for Autonomous Driving
Get PDFSegmentation and recognition become the general steps to identify objects. This research discusses pixel-wise semantic segmentation based on moving objects. The data from the CamVid video which is a collection of autonomous driving images. The image data consist of 701 images accompanied by labels. The segmentation and recognition of 11 objects contained in the image (sky, building, pole, road, pavement, tree, sign-symbol, fence, car, pedestrian and bicyclist) is representing. This moving object segmentation is carried out using SegNet which is one of the Convolutional Neural Network (CNN) methods. Image segmentation on CNN generally consists of two parts: Encoder and Decoder. VGG16 and VGG19 pre-trained networks are used as encoders, while decoders are the upsampling of encoders. Network optimization uses stochastic gradient descent of Momentum (SGDM). The test produces the best recognition was road objects with an accuracy of 0.96013, IoU 0.93745, F1-Score 0.8535 using VGG19 encoder, while when using VGG16 encoder accuracy was 0.94162, IoU 0.92309, and F1-Score 0.8535
Developing an Instrument to Measure Information Assurance Implementation for eGovernment using Goal Question Metric Approach
Get PDFThe eGovernment initiative is aimed to improve government services to the public by improving the quality and availability of services that can be accessed regardless of time and place. Consequently, the services must always be available at any time, and any threat to the information and systems should receive attention to ensure business continuity in the event of an incident. Therefore, in implementing eGovernment, information assurance (IA) must be considered. To determine the extent to which IA implementation status to protect eGovernment services in Indonesia, it is necessary to measure the implementation using an instrument. The measurement instrument was developed using the Goal Question Metric (GQM) approach. The developed instrument was then used in a case study to test its effectiveness in measuring the IA implementation. From the results of the case study, it can be concluded that the IA measurement instrument for eGovernment was proven to be effective within Indonesian context.
 
Diagonal Based Feature Extraction and Backpropagation Neural Network in Handwritten Batak Toba Characters Recognition
Get PDFHandwritten character recognition is considered a complex problem since one’s handwritten character has its characteristics. Data used for this research was a photo of handwritten or scanned handwritten. In this research, Backpropagation Neural Network (BPNN) was used to recognize handwritten Batak Toba character, wherein preprocessing stage feature extraction was done using Diagonal Based Feature Extraction (DBFE) to obtain feature value. Furthermore, the feature value will be used as an input to BPNN. The total number of data used was190 data, where 114 data was used for the training process and another 76 data was used for testing. From the testing process carried out, the accuracy obtained was 87,19 %
Recommended Features of Learning Management System Using Feature-Oriented Method Software Development on Moodle
Get PDFLearning Management System (LMS) serves as a learning media effective for online learning during Covid-19. The rising use of LMS these days indicates that LMS features need to be further measured for their relevance to the National Standards of Higher Education (SN-Dikti). This measurement is also essential to be applied for five blended learning criteria for flexible time and place to access lessons. The objective of this research is to analyze the features of the LMS implemented on Moodle version 3.10. The analysis of the features was carried out by employing the methods of Feature-Oriented Software Development (FOSD), a method that elaborates software system relating to existing features of LMS, involving: 1) Canvas; 2) Edmodo; 3) Google Classroom; 4) Moodle; 5) Zoom Meeting; 6) Google Meet; 7) efront, and 8) Microsoft Teams. In accord with the testing and validation by experts, the research revealed that there were 56 features of LMS pertinent to learning and assessment standards of SN-Dikti and E-learning self-study criteria with other blended-media or events, instructor-led program, live e-learning, on the job training, simulation, and lab centered on blended learning. Those 56 features were approved by experts and implemented by employing Moodle version 3.10 relevant to the SN-Dikti and blended learnin
Early Detection of COVID-19 Patient’s Survavibility Based On The Image Of Lung X-Ray Image Using Deep Neural Networks
Get PDFSARS-CoV-2 causes an infection called COVID-19, which is caused by a new coronavirus. One of the symptomps that dangerous to the patients is developing pneumonia in their lungs. To detect pneumonia symptoms, one of the newest methods is using CNN (Convolution Neural Networks). The problem is when able to detect pneumonia, the patient's survivability, which knowing this will be helpful to decide the priority for each patient, is still in question. The CNN used in this research to classify the patient’s future condition, but met some major problems that the dataset is very few and unbalance. The image augmentation was used to multiply the dataset, and class weight was applied to prevent miscalculation on minority class. 6 CNN architectures used to find the best model. The result VGG19 architecture has the best overall accuracy, in training, it has 80% accuracy, 89% accuracy invalidation, and 82% f1 score accuracy on classifying the testing dataset means the best model if looking for accuracy on prediction, but this cost a prediction time that longest compared to other CNN architectures. MobileNet is the fastest, but it cost much worse on prediction accuracy, only 55%. The ResNet50 model has balanced prediction accuracy/time, it got 77% f1 accuracy, and also 8.49 seconds of prediction time, 9 seconds less than VGG19
Implementation of Oxygen Saturation Sensor Data Acquisition Based on Bluetooth Low Energy Protocol
Get PDFSensor data acquisition is used to obtain sensor data from IoT devices that already provide the required sensor data. To acquire sensor data, we can use Bluetooth Low Energy (BLE) protocol. This data acquisition aims to process further data which will later be sent to the server. Bluetooth Low Energy (BLE) has an architecture consisting of sensors, gateways, and data centers, but with this architecture, there are several weaknesses, namely the failure when sending data to the data center due to not being connected to internet network and data redundancy at the time of data delivery is done. The proposed solution to solve this problem is to create a system that can acquire sensor data using the Bluetooth Low Energy (BLE) protocol with use a store and forward mechanism and checking data redundancy. The proposed system will be implemented using sensors from IoT devices, the gateway used is Android devices, and using the Bluetooth Low Energy protocol to acquire data from sensors. Then the data will be sent to the cloud or server. The results of the test give the results of the system being successfully implemented and IoT devices can be connected to the gateway with a maximum distance of 10 meters. Then when the system stores, for every minute there is an increase in data of 4 kb. Then there is no data redundancy in the system
A Non-Blind Robust and Impercept Watermarking Using Discrete Cosine Transform and Discrete Wavelet Transform
Get PDFNon-blind watermarking is a form of watermarking with a watermark image validation process that requires a host image. The use of the transform domain is more robust and imperceptible. The transform domain method is resistant to various forms of digital image attacks. In this study, Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) were selected as watermark insertion algorithms. DCT is faster and more resistant to attacks, especially in image compression attacks, but has lower imperceptibility than DWT. DWT is also known to be resistant to noise attacks, filtering, blurring, cropping, and has high imperceptibility depending on the sub-band selection but is not resistant to image compression attacks. Based on each algorithm's advantages and disadvantages, there is an opportunity to combine it to analyze and compare the insertion results with DCT and DWT itself. To test the results of imperceptibility, we used the Peak Signal to Noise Ratio (PSNR), while to test the robustness, we used Cross-Correlation (CC) and Bit Error Ratio (BER). Without attacks, the PSNR on the proposed method can reach 71 dB. The CC value without attack can reach a perfect value of 1 and BER = 0. The highest attack test result is CC = 1 on the filtering attack. From the various tests we have conducted, it has been proven that the DCT-DWT is more imperceptive and robust than previous studie
CNN Hyperparameter Optimization using Random Grid Coarse-to-fine Search for Face Classification
Get PDFConvolutional Neural Network (CNN) is a recently used popular machine learning technique to classify images. However, choosing an optimum and efficient architecture is an inevitable challenge. The research goal was to implement CNN on face classification from low quality CCTV footage. The best model was gained from the hyperparameter optimization process used on CNN structure. The optimized hyperparameters were those connected to the structure network including activation function, the number of kernel, the size of kernel, and the number of nodes on the fully connected layers. Hyperparameter optimization strategy used was random grid coarse-to-fine search optimization approach. This approach combined random search, grid search, and coarse-to-fine technique that was easily and efficiently applied, yet worked well. Exhaustive-random search process was done by evaluating all selected activation functions and choosing another hyperparameters randomly. This was based on the assumption that activation functions were the most related hyperparameter to the model. The SELU activation function used in the next step was the one with the best average performance. Grid coarse-to-fine was conducted to optimize the number of kernel and the number of node on fully connected layer, while grid search was conducted to optimize the kernel size. This process aimed to locate optimal value gradually in hyperparameter which had high-dimensional space. Evaluation of the model resulted from the optimum hyperparameter was 97,56%
Design of a Fractional Order PID Controller for Electric Hydraulic Actuator
Get PDFElectric hydraulic actuators are more used especially in industries that demand high levels of accuracy. A common problem with this type of actuator is consistency in fluid flow control. PID controllers can accelerate the achievement of defined output values, eliminate offsets, and reduce maximum overshoots but result in considerable errors. Therefore, it is necessary to design controllers that can reduce errors significantly. In this research, a Fractional Order PID controller is developed to reduce maximum overshoots and steady state. Unlike conventional PID controllers that have three parameters, in the Fractional Order PID controller, there are extra two parameters of the λ and μ. The parameters were selected using the Ziegler Nichols method with a 1st order approach with a delay time. Meanwhile, the λ and μ parameters were selected the best value to make the system response better. The results of the design of the Fractional Order PID controller were evaluated using matlab simulation. The simulation results showed that the Fractional Order PID controller was able to reduce the steady state error response by 0.5 %, and the maximum overshoots by 17.4 %. From this result, it can be noted that the Fractional Order PID controller is better than conventional PID