Journal of Soft Computing Exploration

Activity-based function point complexity of use case diagrams for software effort estimation

Puguh Jayadi — 2024-03-04

This study proposes a Function Point Analysis (FPA) based software development effort estimation methodology integrated with Use Case Diagrams. These methods include identifying actor activities, classifying those activities into FPA categories, and calculating Unadjusted Function Points (UFP). Followed by the calculation of Technical Complexity Factors (TCF) and Adjusted Function Points (AFP), this study aims to produce more accurate man-hours estimates. Results show a UFP of 162 TCF of 11, AFP of 123.12, and an estimated effort of 1846.8 hours worked, while the actual effort is 1228 hours. Evaluation of estimates using the metrics Mean Magnitude of Relative Error (MMER) 0.34, Mean Magnitude of Relative Error (MMRE) 0.50, Mean Absolute Error (MAE) 618.80, Mean Balanced Relative Error (MBRE) 0.50, Mean Inverse Balanced Relative Error (MIBRE) 0.34, and Root Mean Squared Error (RMSE) 618.80, showed sufficient precision despite the overestimation. The study suggests the need for adjustments in TCF calculations and considering development environment factors in more detail to improve estimation accuracy. These findings are essential in improving the precision of effort estimation methodologies in software development, particularly in projects that use Use Case Diagrams as the primary framework.

A sentiment analysis of madura island tourism news using C4.5 algorithm

Vina Angelina Savitri — 2024-03-14

Over the past few years, the tourism sector has experienced significant growth in its contribution. The tourism potential on Madura Island is widespread across four regencies, namely Bangkalan, Sampang, Pamekasan, and Sumenep. This potential can be harnessed to support the local government's economy and the communities in the surrounding areas. This research aims to analyze the sentiment of Madura tourism news from online sources using the Decision Tree (C4.5) method. The data used in this study consist of 100 Madura tourism news articles collected from online news portals, which will be classified using the Decision Tree (C4.5) method. The test results show that this method has an average accuracy rate of 76.5% in 10 tests. The average accuracy results demonstrate that the use of the Decision Tree (C4.5) method in this research yields a sufficiently high accuracy level in the sentiment analysis of tourism news.

Implementation of a reinforcement learning system with deep q network algorithm in the amc dash mark i game

Wargijono Utomo — 2024-03-14

Reinforcement 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.

IoT-based implementation of rickshaws for real-time monitoring and measuring the weight of cattle

Alan Satrya — 2024-03-17

In the era of modern agriculture that is increasingly dependent on technology, livestock management has become crucial to increasing efficiency and productivity. An important aspect in livestock management is providing appropriate feed to fattening cattle. Manual monitoring of feed weight is often complex and prone to errors, which can have a significant impact on operational efficiency and result in losses. Accuracy in monitoring feed weight is the key to maintaining optimal health and growth of cattle. Internet of Things (IoT) technology is emerging as an innovative solution to overcome these challenges. The use of Angkong load cells, a tool connected to IoT, allows automatic monitoring of feed weight with a high level of precision. The test results show an error rate close to zero, with a Mean Absolute Percentage Error (MAPE) of around 0.158%, making the Angkong load cell a reliable tool. With this capability, farmers can monitor cow feed weight in real-time with minimal error rates. This not only increases the operational efficiency of the farm but also optimizes the health and growth of livestock more efficiently, having a positive impact on overall farm productivity. The aim of this research is to monitor the amount of feed given to cows with an adequate level of accuracy. Rickshaw load cells can be well suited for this use due to their ability to handle relatively large weights with fairly good accuracy, but do not necessarily have the level of precision required in laboratory measurements or the pharmaceutical industry.

Prediction of PTIK students' study success in the first year using the c4.5 algorithm

Asri Astuti — 2024-03-17

The purpose of this study is to determine the factors that influence the success of student studies in the first year through data mining research using the C4.5 algorithm. This research is a type of quantitative research. This research uses student data of a study program as much as 85 data which will be processed using the Weka application. The data obtained will then be processed using the C4.5 data mining method to produce a decision tree containing rules to predict the success of student studies in the first year. The best result using percentage-split 80% obtained an accuracy of 82.35% as well as the rules contained in the decision tree. The most important factor in determining the success of studies in first-year students is the selection of college entrance pathways. Other factors that become other determinants are education before college, intensity of communication with friends, class year, intensity of off-campus organizations, and plans to change study programs.

Improved playstore review sentiment classification accuracy with stacking ensemble

Dwi Budi Santoso — 2024-03-18

In today's digital era, user reviews on the Playstore platform are an invaluable source of information for developers, offering insights that are critical for service improvement. Previous research has explored the application of stacking ensemble methods, such as in the context of predicting depression among university students, to enhance prediction accuracy. However, these studies often do not explicitly detail the data acquisition process, leaving a gap in understanding the applicability of these methods to different domains. This research aims to bridge this gap by applying the stacking ensemble approach to improve the accuracy of sentiment classification in Playstore reviews, with a clear exposition of the data collection method. Utilizing Logistic Regression as the meta classifier, this methodology is executed in several stages. Initially, data was collected from user reviews of online loan applications on Google Playstore, ensuring transparency in the data acquisition process. The data is then classified using three basic models: Random Forest, Naive Bayes, and SVM. The outputs of these models serve as inputs to the Logistic Regression meta model. A comparison of each base model output with the meta model was subsequently carried out. The test results on the Playstore review dataset demonstrated an increase in accuracy, precision, recall, and F1 score compared to using a single model, achieving an accuracy of 87.05%, which surpasses Random Forest (85.6%), Naive Bayes (85.55%), and SVM (86.5%). This indicates the effectiveness of the stacking ensemble method in providing deeper and more accurate insights into user sentiment, overcoming the limitations of single models and previous research by explicitly addressing data acquisition methods.

Air quality monitoring using multi node slave IoT

Faisal Fajri Rahani — 2024-03-19

Jakarta is the city with the second poorest air quality in the world. IQAir data show that Jakarta's air quality is 159. In addition, the concentration of air particles in Jakarta is 14.2 times higher than the annual guidelines of the World Health Organization (WHO). According to the WHO, exposure to air pollution causes around 7 million premature deaths and millions of years of lost health time each year. Air pollution also stunts children's growth, impairs lung function, etc. Therefore, we need a system that can be used to combine air quality to determine how dangerous a place is with air quality. Knowing air quality, certain policies or actions being taken to overcome this danger. This research aims to build and test a prototype air quality monitoring system using multi-node slaves with the Internet of Things. The prototype development process was carried out by adapting the architectural framework of the air quality monitoring system with the Internet of Things. The testing of prototype results is carried out to sound sensor values and functional success. The results of the test show that the system can run well according to the design made. The DSM501A sensor device functions to detect particles of a size larger than one micrometer, which usually include cigarette smoke, house dust, ticks, spores, pollen, and mildew, and works well so that the controller can read the surrounding air conditions well.

An optimum hyperparameters of restnet-50 for orchid classification based on convolutional neural network

Nukat Alvian Ideastari — 2024-03-21

There are many types of orchids in Indonesia, such as Phalaenopsis Amabilis (Moon Orchid), Cattleya, etc. Because the shape and color of each orchid flower looks the same, a system is needed that can classify orchid flowers. In this research, we will use a system using a Convolutional Neural Network with ResNet50 architecture to classify orchid species. There are 4 types of orchids that will be used, namely Moon Orchids, xDoritaenopsis Orchids, Cattleya Orchids, and Coelogyne Pandurata Orchids (1000 datasets for each type). The aim of this research is to implement deep learning using the Convolutional Neural Network method combined with the ResNet50 architecture and identifying the types of orchid flowers and calculating accuracy when identifying orchid flower types. This research uses 4000 orchid image datasets, with a data split of 80:20 so that 800 images are used as training data and 200 as test data. ResNet50 uses a confusion matrix evaluation, namely Accuracy, Precision, Recall, Specificity and F1-score with epochs 10, 20, 30, 40. From the research that has been carried out, it produces the highest accuracy on Test Data with the 30th epoch, reaching 98.87%. and the lowest accuracy on Test Data with the 10th epochs which produces an accuracy of 97.75%.

Measuring the usability effectiveness of using card menus and tree menus in school web applications

Hadiq Hadiq — 2024-03-28

The aim of this research is to measure the usability effectiveness of a web application by using card menus and tree menus using user-friendly criteria and access speed as indicated by the number of clicks made by the user. The method used in this research is the Task-centered User Interface method, where this method allows for planning and evaluating the arrangement of the interface according to user needs. There are four stages in this method, including user identification by conducting needs analysis, the second phase is user interface design. The third phase is the implementation of the card menu and tree menu design, and the fourth face is testing the usability and effectiveness requirements. From the research that has been carried out regarding measuring the effectiveness of using card menus, it is more effective to use than tree menus because you can directly lift the menu and access it. Meanwhile, for usability, the card menus have a higher usability index than the tree menus. Meanwhile, for usability measurements carried out by direct observation and distributing questionnaires, the resulting percentage of user understanding, ease, and speed for the card menu display was 87% and for the tree menu was 60% so that the card menu display was more accepted by users than the tree menu. The new thing provided by the results of this research is in the form of suggestions that can be used by web application developers to use the right type of menu in building web-based applications with the same specifications as in the case of school finance applications.

Light sensor optimization based on finger blood estimation and IoT-integrated

Haris Imam Karim Fathurrahman — 2024-03-28

Diabetes mellitus is a prevalent disease in society. This condition results from various causes, such as lifestyle choices or genetic predisposition. To prevent diabetes mellitus, blood glucose levels must be monitored periodically, and dietary consumption must be managed. Blood glucose monitoring still uses the incision or minimally invasive approach. This approach poses a risk of infection and damage. This study devised a method to optimize a light sensor to measure blood glucose levels. This approach uses sensor optimization and an integrated Internet of Things (IoT) technology. The research findings demonstrate that the use of the optimization strategy leads to increased consistency in sensor values, which may then be transmitted wirelessly through the IoT network. The research results demonstrate that using the optimization strategy leads to increased consistency in sensor values, which may then be wirelessly transmitted through the IoT network.

A new CNN model integrated in onion and garlic sorting robot to improve classification accuracy

Apri Dwi Lestari — 2024-04-02

The profit share of the vegetable market, which is quite large in the agricultural industry, needs to be equipped with the ability to classify types of vegetables quickly and accurately. Some vegetables have a similar shape, such as onions and garlic, which can lead to misidentification of these types of vegetables. Through the use of computer vision and machine learning, vegetables, especially onions, can be classified based on the characteristics of shape, size, and color. In classifying shallot and garlic images, the CNN model was developed using 4 convolutional layers, with each layer having a kernel matrix of 2x2 and a total of 914,242 train parameters. The activation function on the convolutional layer uses ReLu and the activation function on the output layer is softmax. Model accuracy on training data is 0.9833 with a loss value of 0.762.

Comparison of gridsearchcv and bayesian hyperparameter optimization in random forest algorithm for diabetes prediction

Rini Muzayanah — 2024-04-02

Diabetes Mellitus (DM) is a chronic disease whose complications have a significant impact on patients and the wider community. In its early stages, diabetes mellitus usually does not cause significant symptoms, but if it is detected too late and not handled properly, it can cause serious health problems. To overcome these problems, diabetes detection is one of the solutions used. In this research, diabetes detection was carried out using Random Forest with gridsearchcv and bayesian hyperparameter optimization. The research was carried out through the stages of study literature, model development using Kaggle Notebook, model testing, and results analysis. This study aims to compare GridSearchCV and Bayesian hyperparameter optimizations, then analyze the advantages and disadvantages of each optimization when applied to diabetes prediction using the Random Forest algorithm. From the research conducted, it was found that GridSearchCV and Bayesian hyperparameter optimization have their own advantages and disadvantages. The GridSearchCV hyperparameter excels in terms of accuracy of 0.74, although it takes longer for 338,416 seconds. On the other hand, Bayesian hyperparameter optimization has a lower accuracy rate than GridSearchCV optimization with a difference of 0.01, which is 0.73 and takes less time than GridSearchCV for 177,085 seconds.

Using genetic algorithm feature selection to optimize XGBoost performance in Australian credit

Dwika Ananda Agustina Pertiwi — 2024-04-03

To reduce credit risk in credit institutions, credit risk management practices need to be implemented so that lending institutions can survive in the long term. Data mining is one of the techniques used for credit risk management. Where data mining can find information patterns from big data using classification techniques with the resulting level of accuracy. This research aims to increase the accuracy of classification algorithms in predicting credit risk by applying genetic algorithms as the best feature selection method. Thus, the most important feature will be used to search for credit risk information. This research applies a classification method using the XGBoost classifier on the Australian credit dataset, then carries out an evaluation by measuring the level of accuracy and AUC. The results show an increase in accuracy of 2.24%, with an accuracy value of 89.93% after optimization using a genetic algorithm. So, through research on genetic algorithm feature selection, we can improve the accuracy performance of the XGBoost algorithm on the Australian credit dataset.