The Effectiveness of Automated Sonic Bloom Method in An IoT-Based Hydroponic System

One of the solutions for food security is planting using hydroponic method and to increase productivity and help hydroponic grow faster and facilitate in monitoring hydroponic growth, sonic bloom and Internet of Things (IoT) are two technologies that can be used. However, in previous studies, the two systems have not been interconnected. The aim of this study is to evaluate the effectiveness of the combination of the two systems mentioned, hence creating an automated sonic bloom method in an IoT-based hydroponic system. To test the proposed method, this system is implemented with bok choi as the hydroponic plant using the DFT technique. The automated sonic bloom is embedded to the IoT system with DF Player Mini module, RTC module, and speakers. The evaluation is done by comparing growth parameters and the crop parameters. The results show that the system with sonic bloom produces fresh weight of 0,44-0,56 g and dry weight of 0,21–0,33 g. The mentioned results are superior to the system without sonic bloom, where fresh weight is 0,17–0,25 g and dry weight is 0,08–0,13 g. It can be concluded that the IoT-based sonic bloom system is effective in increasing the growth rate and hydroponic production rate.

Multivariate Markov Chain Model for Sales Demand Estimation in a Company

Estimation of the number of demands for a product must be done correctly, so that the company can get maximum profit. Therefore, this study discusses how to estimate the amount of sales demand in a company correctly. The model that will be used to estimate sales demand is the Multivariate Markov Chain Model. This model can estimate the future state by observing the present state. The model requires parameter estimation values ​​first, namely the transition probability matrix and the weighted Markov chain, where in previous studies an estimation of the transition probability matrix has been carried out, so that in this study we will continue to estimate the weighted Markov chain parameters. This model is compatible with 5 data sequences (product types) defined as product 1, product 2, product 3, product 4, and product 5, with 6 conditions (no sales volume, very slow-moving, slow-moving, standard, fast moving, and very fast moving). As the result, the state probability for product 1, product 2 and product 3 in company 1 are stationary at state 6 (very fast moving), product 4 and product 5 are stationary at state 2 (very slow moving).

a Schema Extraction of Document-Oriented Database for Data Warehouse

The data warehouse is a very famous solution for analyzing business data from heterogeneous sources. Unfortunately, a data warehouse only can analyze structured data. Whereas, nowadays, thanks to the popularity of social media and the ease of creating data on the web, we are experiencing a flood of unstructured data. Therefore, we need an approach that can "structure" the unstructured data into structured data that can be processed by the data warehouse. To do this, we propose a schema extraction approach using Google Cloud Platform that will create a schema from unstructured data. Based on our experiment, our approach successfully produces a schema from unstructured data. To the best of our knowledge, we are the first in using Google Cloud Platform for extracting a schema. We also prove that our approach helps the database developer to understand the unstructured data better.

The Effect of Number of Factors and Data on Monthly Weather Classification Performance Using Artificial Neural Networks

Current weather-related research only focuses on weather prediction based on raw data and the factors used are generally 4 factors: average temperature, solar radiation, air pressure, and wind. In this research, monthly weather prediction is done using 5 factors where the additional factor used is rainfall in the previous time. In contrast to previous prediction research, the prediction process carried out in this study emphasizes the modeling of training data according to the desired prediction model.. These two things distinguish this research from previous studies. The prediction model used in this study is a classification-based prediction model that is the Artificial Neural Network (ANN) method combined with the backpropagation algorithm for calculating the weight of the ANN network. The data used are meteorological data from 2010 to 2018 in the Bogor area, where data from 2010 to 2016 are used as training data, and data from 2017 to 2018 are used as test data. The results of this study indicate that the design of the model with the use of data for 6 years with feature data of 5 factors has an accuracy rate of 83.33%.

Classification of Hadith Topic of Indonesian Translation Using K-Nearest Neighbor and Chi-Square

Hadith is the main way of life for Muslims besides the Qur'an whose can be applied in everyday life. Hadith also contains all the words or deeds of the Prophet Muhammad which are used as a source of the law of Islam. Therefore, many readers, especially Muslims, are interested in studying hadith. However, the large number of hadiths makes it difficult for readers or those who are still unfamiliar with Islam to read them. Therefore, we conducted a study to classify hadith textually based on the type of teaching, so that readers can get an overview or other reference in reading and searching for hadith based on the type of teaching more easily. This study uses KNN and chi-square methods as feature selection. We also carried out several test scenarios, including implementing stopword removal modifications in preprocessing and experimenting with selecting k values ​​for KNN to determine the best performance. The best performance was obtained by using the value of k = 7 on KNN without implementing chi-square and with stopword removal modification with a hammer loss value of 0.1042 or about 89.58% of the data correctly classified.

STL Decomposition and SARIMA Model: The Case for Estimating Value-at-Risk of Covid-19 Increment Rate in DKI Jakarta

In this research, we propose an extreme values measure, the Value-at-Risk (VaR) based Seasonal Trend Loess (STL) Decomposition and Seasonal Autoregressive Integrated Moving Average (SARIMA) models, which is more sensitive to the seasonality of extreme value than the conventional VaR. We consider the problem of the seasonality and extreme value for increment rate of Covid-19 forecasting. For stakeholder, government and regulator, VaR estimation can be implemented to face the extreme wave of new positive Covid-19 in the future and minimize the losses that possibly affected in term of financial and human resources. Specifically, the estimation of VaR is developed with the difference lies on parameter estimators of STL and SARIMA model. The VaR has coverage probability as well as close 1-α. Thus, we propose to set α as parameter to estimate VaR. Consequently, the performance of VaR will depend not only on parameter model but also α. Our aim estimates VaR with minimum α based on correct VaR value. Numerical analysis is carried out to illustrate the estimative VaR.

Toxic Comment Classification on Social Media Using Support Vector Machine and Chi Square Feature Selection

The use of social media in society continues to increase over time and the ease of access and familiarity of social media then make it easier for an irresponsible user to do unethical things such as spreading hatred, defamation, radicalism, pornography so on. Although there are regulations that govern all the activities on social media. However, the regulations are still not working effectively. In this study, we conducted a classification of toxic comments containing unethical matters using the SVM method with TF-IDF as the feature extraction and Chi Square as the feature selection. The best performance result based on the experiment that has been carried out is by using the SVM model with a linear kernel, without implementing Chi Square, and using stemming and stopwords removal with the F1 − Score equal to 76.57%.

Neural Network on Stock Prediction using the Stock Prices Feature and Indonesian Financial News Titles

Stocks are the most popular investments among entrepreneurs or other investors. When investing in stocks these investors tend to learn how to invest stocks correctly and when is the right time. For the problem of how to invest shares correctly can be used a variety of basic theories that already exist, but for the problem when the right time needs further learning. In this paper will purpose about stock price prediction using stock data indicators and financial headline data in Bahasa Indonesia. The machine learning model used is a multi-layer perceptron neural network (MLP-NN) with the highest accuracy produced by 80%.

Performance Comparison of Several Range-based Techniques for Indoor Localization Based on RSSI

The classical rang-based technique for position estimation is still reliably used for indoor localization. Trilateration and multilateration, which include three or more references to locate the indoor object, are two common examples. These techniques use at least three intersection-locations of the references' distance and conclude that the intersection is the object's position. However, some challenges have appeared when using a simple power-to-distance parameter, i.e., received signal strength indicator (RSSI). RSSI is known for its fluctuated values when used as the localization parameter. The improvement of classical range-based has been proposed, namely min-max and iRingLA algorithms. These algorithms or methods use the approximation in a bounding-box and rings for min-max and iRingLA, respectively. This paper discusses the comparison performance of min-max and iRingLA with multilateration as the classical method. We found that min-max gives the best performance, and in some positions, iRingLA gives the best accuracy error. Hence, the approximation method can be promising for indoor localization, especially when using a simple and straightforward RSSI parameter.

Forecasting the COVID-19 Increment Rate in DKI Jakarta Using Non-Robust STL Decomposition and SARIMA Model

The Corona virus known as COVID-19 was first present in Wuhan, China at this time has troubled many countries and its spread is very fast and wide. Data on daily confirmed COVID-19 cases were collected from the DKI Jakarta province between early May 2020 and late January 2021. The daily increase in confirmed COVID-19 cases has a percentage of the value of increase in total cases. In this study, modeling and analysis of forecasting the increment rate in daily number of new cases COVID-19 DKI Jakarta was carried out using the Seasonal-Trend Loess (STL) Decomposition and Seasonal Autoregressive Integrated Moving Average (SARIMA) models. STL Decomposition is a form of algorithm developed to help decompose a Time Series, and techniques considering seasonal and non-stationary observation. The results of the best forecasting accuracy are proven by STL-ARIMA, there are MAPE and MSE which only have an error value of 0.15. This proposed approach can be used for consideration for the DKI Jakarta government in making policies for handling COVID-19, as well as for the public to adhere to health protocols.