Search

Technology development has been getting more advanced and greatly facilitated human life. One of them is machine learning automation which has been proven to be consistent for doing various computations against extensive data such as transaction data in the e-commerce area. Seeing this opportunity, we implemented the machine learning approach to detect fraudulent cashback transactions in e-commerce that are currently rife in Indonesia. The training data used to build the machine learning model were the transaction data from one of the leading e-commerce in Indonesia that had been processed. The supervised classification algorithms used were K-Nearest Neighbor (k-NN), Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM). In the end, the best steps and methods that could be taken against fraudulent cashback activities in the future are shown in this paper. [ABSTRACT FROM AUTHOR]

Nowadays, social media is considered as an essential tool for an organization to communicate with another user, customers and potential customers for gaining popularities, getting ideas and opinion from the users, or affecting another user. One of the problems on social media is the uses of the social bots, that used to make a human user believe the opinion generated from the social bots. These bots can use their ability to post in higher rates than human user to shift the topic that discussed in the time being. To resolve the problems on social bots, there are several ways to identify a bot from an account such as checking the date of account creation and interval on creating between posts with using machine learning.in this paper we will discuss what is the best machine learning algorithm to determine a social bot, and which feature that helps the algorithm the most. In our work, the data that will be used from Indonesia President Candidate's Supporter's Tweets from each side, Joko Widodo (@jokowi) as the first candidate and Prabowo Subianto (@prabowo) as the second candidate from a month. The algorithms that will be used on this paper are Support Vector Machine and Random Forest with cross-validation as a training method. We found that the year of account created has the highest contribution to bot detection. [ABSTRACT FROM AUTHOR]

The sentiment classification method is a research field that is proliferating in Indonesia since it is fast in extracting public opinion and provides essential and valuable information for stakeholders. Of the best-performing sentiment classification approaches, machine learning is one of them that has excellent performance. However, the method has several problems, such as noisy features and high dimensionality of features that significantly affect the sentiment classification performance. Therefore, to overcome the problems, this paper presents a novel feature selection using a combination of Query Expansion Ranking (QER) and Genetic Algorithm-Support Vector Machine (GA-SVM) for improving sentiment classification performance. Based on the experimental results, the proposed method could significantly improve sentiment classification performance, outperform all state-of-the-art algorithms, and decrease computational time. The method achieved the best performance in average precision, recall, and f-measure with the value of 96.78%, 96.76%, and 96.75%, respectively. [ABSTRACT FROM AUTHOR]

Today, Indonesian citizen experience a fast changing lifestyle related to information and communication technology (ICT) utility, such as gadget and internet. The ease of information and knowledge retrieval to support learning process comes with problems to be taken care of such as information overload, negative content, netiquette negligence, and gadget addiction. Meanwhile, Indonesia will face an upcoming challenge in the economic sector, Free Trade Area of the Asia-Pacific (FTAAP) in 2020. An initiative for economic integration uses ICT as economic hub to promote every sector growth in economy. A concept explored in this research context to answers those problems mentioned above and to overcome the upcoming challenge is improving quality of Indonesian human resources through digital literacy education. This paper discusses a learning system model dedicated to deliver digital literacy education to the Indonesian citizen. Objectives of this research are increasing awareness towards digital literacy concept and preparing citizen to be digitally literate and competent using ICT to support their learning process, doing their work more effective and efficient, and preparing Indonesian young generation to become qualified workforce competing in FTAAP 2020, furthermore international work competition. This paper delivers analysis for learning system model as a result from the initial research of research series about digital literacy education in Indonesia. Resulting model justified by sets of detailed scenario about how this model works. [ABSTRACT FROM AUTHOR]

Dengue fever is a disease caused by the dengue virus, which is spread by Aedes aegypti and Aedes albopictus mosquitoes. According to the WHO, as a tropical country, Indonesia is a country at high risk for dengue. Dengue can spread to other people through mosquito bites. Weather factors, such as temperature, humidity, and rainfall, affect the number of dengue incidents. It is important to predict the number of dengue incidents so that the government and the people will be ready to prevent a dengue outbreak when the number of dengue incidents is predicted to be high. In this paper, we predict the number of dengue incidents in DKI Jakarta using support vector regression, with weather and the previous number of incidents as predictor variables. These predictor variables are determined by analyzing the time lag between each predictor variable and the number of incidents by using cross-correlation. Models for prediction are compared by Root Mean Squared Error and Mean Absolute Error. The result shows that support vector regression with linear kernel is quite good, and is in fact better than the radial kernel, for predicting the number of dengue incidents. [ABSTRACT FROM AUTHOR]

Recently, horticulture farming, especially in Indonesia, are facing various obstacles, including the lack of young farmers regeneration, agricultural land decreasing, and climate change. Climate change increase extreme and uncertain weather for horticulture cultivation. Farmers frequently face frequent damage to crops and decreasing in quantity and quality of plants. The development of a monitoring system that can help farmers grow crops is a crucial issue. In this paper, a study for a system for monitoring the performance of indoor micro-climate horticulture is presented. An electronic sensors Internet-of-Things (IoT) board is implemented to monitor the cultivation process. The sensor reading is collected into a database. The database is also integrated with the Indonesian Meteorological Agency data which consists of weather data and daily at the cultivation location. For our future work, the collected data are expected to be used for training machine learning models for the development of intelligence automated indoor micro-climate horticulture. [ABSTRACT FROM AUTHOR]

Recently there has been an interest in predicting complex disease risk using models that combine the effects of many genetic factors together. These are known as polygenic models and are useful in evaluating risk of disease in patients. These models, however, often do not include important non-genetic factors that are important to the prediction of the disease. In this paper, we explore the prediction of colorectal cancer in Indonesia from non-genetic factors using common machine learning algorithms: XGBoost and Elastic Net. The result of this study identified 8 features with strong importance from both XGBoost and Elastic Net. We recommend including these features as covariates in future genetic association studies of colorectal cancer in Indonesia. Ultimately these models may be implemented as tools to screen patients for colorectal cancer risk. [ABSTRACT FROM AUTHOR]