Tessy Badriyah

High-dimensional data allows researchers to conduct comprehensive analyses. However, such data often exhibits characteristics like small sample sizes, class imbalance, and high complexity, posing challenges for classification. One approach employed to tackle high-dimensional data is feature selection. This study uses the Bacterial Foraging Optimization (BFO) algorithm for feature selection. A dedicated BFO Java library is developed to extend the capabilities of WEKA for feature selection purposes. Experimental results confirm the successful integration of BFO. The outcomes of BFO's feature selection are then compared against those of other evolutionary algorithms, namely Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), and Ant Colony Optimization (ACO). Comparison of algorithms conducted using the same datasets. The experimental results indicate that BFO effectively reduces features while maintaining consistent accuracy. In 4 out of 9 datasets, BFO outperforms other algorithms, showcasing superior processing time performance in 6 datasets. BFO is a favorable choice for selecting features in high-dimensional datasets, providing consistent accuracy and effective processing. The optimal fraction of features in the Ovarian Cancer dataset signifies that the dataset retains a minimal number of selected attributes. Consequently, the learning process gains speed due to the reduced feature set. Remarkably, accuracy substantially increased, rising from 0.868 before feature selection to 0.886 after feature selection. The classification processing time has also been significantly shortened, completing the task in just 0.3 seconds, marking a remarkable improvement from the previous 56.8 seconds.

Shrimp farming plays a crucial role to the Indonesian economy, but it is facing challenges from shifting weather patterns and global warming. This research focuses on the development and implementation of a web-based water quality monitoring system for shrimp farming to address these concerns. The research, conducted in collaboration with shrimp farmers in Sidoarjo, East Java, introduces PENS Aquaculture program, which is designed to efficiently monitor pH, salinity, and temperature. The system employs Internet ofThings (IoT) technology, which allows farmers to register several ponds, analyze water parameters, and receive real-time data through tables and graphs. The research takes a mixed-methods approach, integrating quantitative data from IoT devices with qualitative insights gathered through surveys and interviews with shrimp farmers. The study aims to evaluate the influence of IoT technology on shrimp pond quality and its contribution to the production. The findings show that PENS Aquaculture application is helpful in increasing shrimp farming efficiency, providing significant insights for the fisheries and cultural sectors.

Stroke may cause death for anyone, including youngsters. One of the early stroke detection techniques is a Computerized Tomography (CT) scan. This research aimed to optimize hyperparameter in Deep Learning, Random Search and Bayesian Optimization for determining the right hyperparameter. The CT scan images were processed by scaling, grayscale, smoothing, thresholding, and morphological operation. Then, the images feature was extracted by the Gray Level Co-occurrence Matrix (GLCM). This research was performed a feature selection to select relevant features for reducing computing expenses, while deep learning based on hyperparameter setting was used to the data classification process. The experiment results showed that the Random Search had the best accuracy, while Bayesian Optimization excelled in optimization time.

Development of technology causes many business industries to migrate from offline business systems to the e-commerce world. One of the most popular e-commerce frequented by potential buyers is the property site. Considering that the property is one of the essential requirements for living, and furthermore it is also one of the most prized assets one can have. In this research, we develop a web-based recommendation system in choosing a property using content-based filtering method. The recommendation system provides property information based on user behavior by searching advertising content previously searched by the user. Each time the user selects the contents of the ad to display, this information will be stored into the database to be processed further in order to provide a recommendation. The application system will present the same product recommendation, in accordance to the profile / criteria and preference of the prospective buyer. Therefore, the recommendation system will assist prospective buyers in determining the choice of property product they want to buy, and this process can be provided by the recommendation system in a short time.

The number of cases of cancer in Indonesia is very high. This can be seen from a variety of data published by the government or any cancer research institutes. Cervical cancer has been named as the number one killer of women in Indonesia, followed by breast cancer with an estimated number of 61,682 cases in 2013. This is despite the fact that cancer is a disease that can be cured when detected early, as claimed by an Oncology doctor. Therefore, came the idea of this research with the aim to provide predictive tools of a decision support to assess the level of malignancy of a tumor with a case of data Oncology Hospital Surabaya. In this research, we develop mobile personal health record (mPHR) for Breast Cancer disease using methods in Machine Learning as predictive tools. In this case, Logistic Regression was compared with Naive Bayes method used to predict the risk of of tumor malignancy related to Breast Cancer disease.

Most of cardiovascular disorders or diseases can be prevented, but death continues to rise due to improper treatment because of misdiagnose. One of cardiovascular diseases is Arrhythmia. It is sometimes difficult to observe electrocardiogram (ECG) recording for Arrhythmia detection. Therefore, it needs a good learning method to be applied in the computer as a way to help the detection of Arrhythmia. There is a powerful approach in Machine Learning, named Deep Learning. It starts to be widely used for Speech Recognition, Bioinformatics, Computer Vision, and many others. This research used the Deep Learning to classify the Arrhythmia data. We compared the result to other popular machine learning algorithm, such as Naive Bayes, K-Nearest Neighbor, Artificial Neural Network, and Support Vector Machine. Our experiment showed that Deep Learning algorithm achieved the best accuracy, which was 76,51%.

E-commerce is an online trading system that eases transactions for both sellers and consumers without having to meet in person. The prevalence of e-commerce has increased competition amongst sellers, hence the users of e-commerce has to increase their performance, one of them by using recommendation system. This research develops a hybrid recommendation system for e-commerce that implements Content-based Filtering and Collaborative Filtering methods, which will compute the simmilarities of product description and user profile. In experiment results, it was found that the recommendation has similarity with product description and the preference of user profile with the average of precision value is 67.5% and recall value is 71.47%.

Wireless sensor networks (WSNs) have been widely used for monitoring, tracking or object navigating whether stationer or mobile object. The system could be implemented in indoor or outdoor area. Disaster activity in urban area such as earthquake, tornado even militant activity inside the building need an object tracking in order to carry out search and rescue procedure for the victims or rescuers. Localization technique in indoor area encounters some problems which related to environment characteristics such as building material, multipath, reflection and other parameters which will decrease the accuracy of estimated object position. We propose a geometric triangulation scheme as an alternative way of calculating the estimated position of the object, based on the angle of arrival between an object node and its references. This scheme is free from environment problem. The simulation result shows that using least square method through the average of position error the estimated position accuracy will be increased up to 22.27% better than the previous triangulation scheme.

The placing components and routing are very important in designing PCB, although they require a lot of time and precision. Auto placer and auto router application that are available so far are proprietary and thus cannot be developed freely. This research proposes optimization design of PCB alongside with genetic algorithm and routing lee algorithm. The genetic algorithm allocates components automatically and Routing Lee algorithm does routes between components.