SHM Publisher Journals
Not a member yet
315 research outputs found
Sort by
Implementation of a reinforcement learning system with deep q network algorithm in the amc dash mark i game
Get PDFReinforcement learning is a branch of artificial intelligence that trains algorithms using a trial-and-error system. Reinforcement learning interacts with its environment and observes the consequences of its actions in response to rewards or punishments received. Reinforcement Learning uses information from every interaction with its environment to update its knowledge. The problem identified from this research is the lack of consistency, which is not always the same for Non-Player Characters (Agents) in the process of exploring an environment (Game environment). This research uses the Software Development Life Cycle (SDLC) Waterfall model method to train Non Player Characters (Agents) in the Amc Dash Mark I Game which uses the Deep Q Network (DQN) algorithm in several stages. Training results show improvements in model performance over time. The average duration of the episode and average reward episode showed an increase of 7.75 to 24.7, while the exploration rate decreased to 0.05. This indicates that the model has experienced learning and is improving to achieve better rewards by performing fewer actions. The lower loss also shows that the model has succeeded in reducing prediction errors and improving prediction capabilities
Enhancing soccer pass receiver prediction in broadcast images through advanced deep learning techniques: A comprehensive study on model optimization and performance evaluation
Get PDFIn this study, we present a graph neural network (GNN) model specifically designed for football pass receiver prediction in Broadcast Images is presented in this study. Important node properties, including ball possession indicators, hot-encoded team values, and normalized ground placements, are incorporated into the model along with a careful weighting of edges to account for player distances. With weighted BCE loss used to overcome class imbalance, its architecture consists of a linear layer, numerous GNN Message Passing layers, a SoftMax activation, and binary cross-entropy (BCE) loss for training. Of 206 examples, 101 valid predictions were made, indicating a predictive accuracy of 0.50 according to the evaluation data. Comparative analyzes show that GAT-V2 (0.85) and GAT (0.63) perform better in terms of optimization and accuracy, respectively. The effectiveness in recognizing football pass receivers is demonstrated in this paper, highlighting developments in computer vision applications for sports analytics
Brain Computer Interface (BCI) Machine Learning Process: A Review
Get PDFThe abstraction of Brain Computer Interface (BCI) is a communication and control system that translated human mind thoughts into real-world interaction without any use of neural pathways and muscles. BCI is used as a tool to help person that suffered from impairment to be able do their daily activities independently. In general, BCI process its signal through several process such as pre-processing, and classification. However, providing information of pre-processing and classification process is barely found. Therefore, in this review paper we present the various pre-processing and classification methods that used in the BCI system application.
 
Identification of tourism potential and 3A analysis (attraction, amenity, accessibility) in maitara village of akebay village
Get PDFThis research aims to identify tourism potential and analyze 3 A (attraction, accessibility and amenties) in Maitara Village of Akebay Village. It is hoped that from this research the government and the community of Maitara Kampung Akebay Village will get an overview and guidelines on tourism development in Akebay Village. Tourism development certainly requires attractions to be used as tourist attractions, easy access such as roads and transportation and additional supporting facilities such as lodging, restaurants, toilets, and so on which are called the 3 A terms.
From the results of field research there are several potentials that can be used as tourist attractions in Maitara Village, Akebay Village including, Nature Tourism, Culture and Fitness Tourism. The theory used is the theory of potential, tourist attraction and 3A.
The conclusion of this research is that there are several types of attractions that can be offered by Maitara Kampung Akebay Village as a tourist attraction, namely (1). Nature tourism consisting of Camping Ground, Tracking, (2) Cultural Tourism consists of "Bakerah" and traditional games, (3) Artificial Tourism consists of intagramable photo spots and swimming pools for children
The influence of employee service on customer satisfaction at café V2 coffee space Tembalang
Get PDFThis research was conducted to analyze guest comments and determine customer satisfaction at V2 Coffee Space Tembalang Cafe. The method used in this research is descriptive, while the method of presenting the data is quantitative by taking the research site V2 Coffee Space Tembalang Cafe. The data collection techniques are observation, guest comment and literature study. The results of data analysis show that customers who come to Cafe V2 Coffee Space Tembalang with 50 respondents stated as follows: 1) Customer responses about guest comments on Food quality can be said to be good as seen in the results of the analysis, 7 people stated Excellent, 32 people stated Very Good, 1 person stated Fair. 2) Customer responses about Guest Comment on menu Variety can be said to be good as seen in the analysis results, 10 people stated Excellent, 30 people stated Very Good, 10 people stated Fair. 3) Customer responses about Guest Comment on Food Quality can be said to be quite good as seen in the analysis results, 14 people stated Excellent, 25 people stated Very Good, 5 people stated Fair, 6 people stated Poor. 4) Customer responses about guest comment on Food Quality can be said to be good as seen in the results of the analysis, 21 people stated Excellent, 28 people stated Very Good, 1 person stated Fair
Deep convolutional generative adversarial networks for data imbalance in convolutional neural networks for facial expression classification
Get PDFFacial expression recognition technology is a critical direction of emotion computing research, and it is an essential part of human-computer interaction. The facial expression recognition method is a classification method. An excellent classification method and widely used today are the Convolutional Neural Network (CNN). However, there are still shortcomings in accuracy in the CNN method if the available dataset is minimal and imbalanced. There are two ways to overcome this, adding the training data or changing the architecture on CNN. In this research, the researcher uses the method to add to the training dataset using the Deep Convolutional Generative Adversarial Networks (DCGAN) method
The Effect of Digitalization on Business Performance in the MSME Industry Context
Get PDFThe current digital era is increasingly developing in the use of new technology that creates value for companies and offers benefits. Digitalization is useful for increasing competitive advantage to improve business performance. The purpose of this study is to find out whether digitalization affects business performance and to find out whether competitive advantage can mediate digitalization on business performance. The sample of this research is 115 SMEs in Semarang. data were analyzed using the SEM approach with the smartPLS tool. The results of the study show that the digitalization variable has an influence on business performance, furthermore, competitive advantage also has a positive and significant effect on business performance. The results of the indirect effect test also show that competitive advantage can mediate the relationship between digitalization and business performance. The better the implementation of digitalization, the higher the competitive advantage MSMEs, consequently leading to an increase the business performance
The Optimization of Credit Scoring Model Using Stacking Ensemble Learning and Oversampling Techniques
Get PDFCredit risk assessment plays an important role in efficient and safe banking decision-making. Many studies have been conducted to analyze credit scoring with a focus on achieving high accuracy. However, predicting credit scoring decisions also requires model construction that handles class imbalance and proper model implementation. This research aims to increase the accuracy of credit assessment by balancing data using Synthetic Minority Oversampling (SMOTE) and applying ensemble stacking learning techniques. The proposed model utilizes a base learner consisting of Random Forest, SVM, Extra-Tree Classifier, and XGboost as a meta-learner. Then to handle unbalanced classes using SMOTE. The research process was carried out in several stages, namely Data Collection, Preprocessing, Oversampling, Modeling, and Evaluation. The model was tested using the German Credit dataset by applying cross-validation. The evaluation results show that the stacking ensemble learning model developed has optimal performance, with an accuracy of 83.21%, precision of 79.29%, recall of 91.78%, and f1-score of 85.08%. This research shows that optimizing the stacking ensemble learning model with data balancing using SMOTE in credit scoring can improve performance in credit scoring
Potential of purple eggplant skin fraction (Solanum Melongena Var. Serpentinum L.) as an in vitro sunscreen
No full textSunscreen is a product that is used as skin protection against sun exposure. The skin of purple eggplant is known to contain secondary metabolites which are used as sunscreens. This study aims to determine the activity of sunscreen fraction of the purple eggplant (Solanum melongena var. serpentinum L.) in vitro. Testing of sunscreen activity was carried out by calculating the value of SPF (Sun Protection Factor), %TE (Percent Transmission of Erythema) and %TP (Percent Transmission of Pigmentation) using the UV-Vis Spectrophotometry method. The purple eggplant skin fractionation method is by using LVC (Liquid Vacuum Chromatography). The results of the fractionation were identified using phytochemical screening and combined TLC. The fraction results were tested for SPF (Sun Protection Factor), %TE and %TP values. The results showed that the extract yield was 26.33%. The fractions used are fraction 9; 10; 11; and 12 which showed the same stain after being evaporated, which was a brownish yellow stain with a yield of 21,24%. The purple eggplant skin fraction (Solanum melongena var. serpentinum L.) has SPF, %TE and %TP values with optimal protection at a concentration of 400 ppm with each value obtained being 19; 1.35 and 9.79. The results of the statistical test of the purple eggplant skin fraction showed that the data were normally distributed but not homogeneous. The results of the homogeneity test for each concentration of results showed that the data were not significantly different
Early Detection of Microsleep in Motorcycle Helmet Based on Pulse Sensor
Get PDFMicrosleep can be defined as a brief condition in which someone unintentionally falls asleep for a few seconds to several minutes. This condition can occur in anyone and poses a high potential risk, especially when engaged in activities that require high concentration, such as driving. To detect and address the potential dangers of microsleep while driving, this research has designed a smart helmet capable of early detection of signs of microsleep and taking actions to awaken the rider. This system uses a pulse sensor connected to an Arduino and placed on the backside of the helmet. Detection of beats per minute (bpm) is crucial to determine whether the rider is drowsy or not. This is essential for providing early warnings to the rider. If the rider's bpm reading is <60, indicating drowsiness, the system activates a vibrator to shake the helmet. If this condition persists for more than 7 seconds, the speaker also activates to play music, which will only stop when the bpm reading is >60. Testing was conducted on 5 test subjects, with each subject undergoing 5 trial tests, resulting in a total of 25 test runs. The results indicate that the designed system is capable of reading microsleep conditions and activating the vibrator and music according to the configured settings