Publication Search

Tabular-based clinical risk prediction models are extensively applied in medical decision support systems; however, two major challenges often reduce their reliability: predictions that contradict basic clinical logic and poorly calibrated probability outputs that weaken threshold-based decision making. This study investigates explainable binary risk prediction using the processed Cleveland subset of the UCI Heart Disease dataset as a public clinical benchmark. A lightweight and CPU-efficient pipeline is proposed by employing an XGBoost classifier integrated with monotonic constraints on clinically relevant features, followed by probability calibration through post-hoc methods, including Platt scaling, temperature scaling, and isotonic regression on a separate validation set. Model performance is assessed in terms of discrimination capability using AUROC, AUPRC, F1-score, sensitivity, and specificity, while probability reliability is evaluated using ECE and Brier score metrics. A monotonicity audit is also conducted through counterfactual feature sweeps to measure violation rates. In addition, the model is applied for risk stratification into low-, medium-, and high-risk categories with corresponding event-rate reporting. The findings demonstrate that isotonic regression improves probability reliability without degrading discrimination performance. Furthermore, the monotonicity audit reveals no observed violations for constrained features. Overall, the integration of monotonic constraints and probability calibration produces more decision-ready risk estimates for threshold-based clinical decision support while maintaining transparency through SHAP-based analysis.

Consensus problems need communication between two or more agents. The existence of time delay in communication makes every agent doesn’t get the real-time states of the other agents. The main problem of delay system is the response starts slowing down and oscillating when the gain is increasing. This paper proposes a predictor-feedback that reduces the effect of time delay. The predictor itself utilizes the complete subgraphs. Analytically the result generates faster response compared to the system without the predictor. Then, the proved solution is supported by a numerical solution.

Technological developments today make it easy for people to use social media as a means of expressing opinions, including Twitter. The case study taken by the researcher is the sentiment towards the performance of the mayor of Medan. The case was taken because it was widely discussed by Indonesian people, especially the city of Medan on Twitter social media. One of the uses of this research is to find out the trend of Twitter user comments on the performance of the mayor of Medan by conducting a sentiment analysis. Sentiment will be classified as positive, negative and neutral. The algorithm used in sentiment analysis is Naïve Bayes. The stages in conducting sentiment analysis in this study are data preprocessing, data processing, classification, and evaluation. The results of this study are using the SMOTE method, the training and testing ratio is 80:20 because it has the highest accuracy, which is 78% compared to other ratios. The prediction results resulting from the classification turned out to be more dominant towards neutral labels. In addition to classifying for sentiment analysis, this study also measures the performance of the model created. The results showed that the Naïve Bayes algorithm has a precision value of 78%, a recall of 78%, and an f1-score of 77%.

Property price determination is a complex challenge influenced by various factors, thus requiring an effective method for accurate prediction to support investment decision-making. In the current digital era, conventional approaches are being replaced by data-driven and artificial intelligence methods, where Linear Regression remains a popular choice due to its simplicity and effectiveness in modeling linear relationships. This study aims to analyze the relationship between the physical characteristics of a house and its selling price, and to build an accurate predictive model using the Linear Regression algorithm. A quantitative method was used, focusing on Building Area , Number of Rooms, and Building Age  against the House Selling Price. Correlation analysis results show that Building Area has the strongest correlation (0.81) with price, while Building Age shows a negative correlation (-0.52). The Linear Regression model demonstrated very strong and stable performance. The model achieved an R² Score of 0.9396 on the testing data, meaning 93.96% of house price variability can be explained by the model. Furthermore, the low MAE of only 11.31 million rupiah indicates a small prediction error, and the consistency of R² scores confirms that the model does not suffer from overfitting. This study concludes that the Linear Regression model provides excellent, stable, and reliable prediction performance for projecting house selling prices

Perum Bulog as a State-Owned Enterprise has the main task, which is to conduct a quality and adequate basic food logistics business for the survival of the people. The problem that occurred in Perum Bulog Sub Divre Medan is that the rice supply in Bulog does not consider the demand in the market. Forecasting is an important tool in effective and efficient planning. Therefore, prediction is indispensable for predicting future events. This method essentially uses past data initiated by performing an exponentially decreasing weighting of older observational values or newer values. Brown's double exponential smoothing is a linear model proposed by Brown. This double exponential smoothing method is used when the data indicate a trend. In this study, the terbaik best parameter for forecasting the Number of Rice Sales in Perum Bulog Sub Divre Medan was α = 0.2 with MAPE of 0.27%. And the results of the forecast for Rice Sales at Perum Bulog Sub Divre Medan in 2022 are decreasing every month.

In developing a business or sale is to follow technological developments including the use of systems for buying and selling interactions. There are already many sellers who make buying and selling interactions online. In addition, to develop a business, it is also necessary to predict future sales so that the seller knows and prepares the number of goods to be sold to avoid shortages or excess quantities of goods. To find sales predictions, various methods can be used, one of which is Double Exponential Smoothing method. Double Exponential Smoothing  method is the time series method that uses data from the past to predict the next period. The data processed is sales data at Dasni clothing stores for one year. The results obtained are in the form of a sales prediction system for the next 3 months period which calculates the level of prediction accuracy using MAPE (Mean Absolute Percentage Error) with the smallest error sought because the smaller the error, the more accurate it is to predict the number of sales in the next period. This prediction system is also designed using the PHP programming language.

The financial system plays an important role in the economy. An unstable financial system will be vulnerable to various problems that disrupt the rotation of a country's economy and be vulnerable to economic problems such as the global crisis in various countries. The problem that occurs is the occurrence of Covid-19 causing various fluctuations in the level of inflation, money supply, imports, the occurrence of unstable inflation from January 2019 to August 2021, low inflation resulting in a decrease in imports and an increase in the money supply in Mexico. , Vietnam, Philippines, Hongkong, Indonesia, Canada, Malaysia, Singapore, Peru, and China. The analytical method in this study uses the ARDL Panel (Autoregression Distributed Lag) approach. The ARDL Panel Model determines which country models from APEC countries are able to control long-term financial system-based economic fundamentals in Mexico, Vietnam, the Philippines, Hong Kong, Indonesia, Canada, Malaysia, Singapore, Peru, and China and the Different Test for modeling the impact of covid-19 19 on the economic fundamentals of the financial system. The results of the research found the ARDL Panel prediction model in modeling the impact of Covid-19 on economic fundamentals in the financial system. The main Leading Indicator of variable effectiveness in controlling Inflation In TAPEC is JUB where Vietnam, the Philippines, Hong Kong, Japan, Malaysia, Singapore, Peru and China have a significant influence in controlling Inflation. Then overall in the long term (Long Run) it turns out that only the JUB and CDV variables have an effect on INF In TAPEC, while in the short term (Short Run) it is JUB that influences Inflation In TAPEC.  

Accurate weather prediction information is important for various fields that are closely related to weather forecasting, such as agriculture, fisheries and many more. Because precise weather forecasts are very useful for various fields of carrying out various activities. Because of that, it is necessary to make an application to find weather or rainfall prediction information, so that the information can be utilized optimally by the community. In this journal the authors apply the k-nearest neighbors (k-NN) method based on rainfall data obtained from the Bogor climatology station from 2016-2017 and the test results show that the predicted rainfall for the Bogor area with the K-Nearest Neighbor algorithm obtained a value of 0, 93148.  

The purpose of this study is to produce a design to predict, analyze and determine the level of potential bankruptcy of a company using the Altman Z-Score method. Predictions are made by analyzing the financial statements of a company. The research approach used is a qualitative approach. Data analysis technique in this research is descriptive analysis technique. The results of the first phase of research are in the form of a review of bankruptcy prediction analysis using secondary data from banks in Indonesia that are already on the stock exchange and have branch offices in the city of Semarang, while the second year produces an information system design that has added value from the first year to the third year. testing the system that has been designed using actual financial statement data.

Management of medical actions carried out in handling patients who are ODP (people under monitoring), OTG (asymptomatic people), PDP (patient under monitoring) and positive Covid-19 patients is carried out based on assumptions, such as self-isolation, hospitalization, or special treatments in the ICU (Intensive Care Unit) room. The condition of the body in each patient is different, a patient may have same symptoms but the treatment is different, especially in elderly patients. Many problems occur in determining medical action because the patient's body condition is different. Therefore, it needs to be appointed as a research. The research method used in this study was Nive Bayes algorithm with supporting application Rapid Miner. It was applied to carry out the process of testing on patient data as much as 500 data, 25 variables or patient symptoms and 3 outputs as a form of medical action. Based on the results of the analysis carried out in this study, prediction of medical actions for ODP, PDP, OTG and positive Covid-19 patients were obtained by comparing training data with testing data using Rapid Miner application. It resulted that an accuracy rate of 76.00% was obtained

The AWD Mranggen store is a store that is engaged in the sale of bags, belts, shoes with sales developments increasing from year to year, with fairly tight business competition, the AWD Mranggen store must be able to calculate the estimated number of items to be purchased based on previous sales data, the prediction is very influential on the decision to determine the number of items to be provided by the AWD Mranggen Store for the next sales period data. Inventory of goods that are not right cause some losses in terms of time and also costs, it is necessary to have a forecasting system. Forecasting is a technique to identify a model that can be used to predict conditions in the future. By using the weight moving average method, it can be seen that the error value is more than smaller than other methods and the estimated results can be more precise so that it can help owners make decisions in carrying out inventory.

Chronic kidney failure is the failure of kidney function in maintaining metabolism and fluid and electrolyte balance in the body. Chronic kidney disease initially does not show significant symptoms and signs but can develop rapidly into kidney failure. Kidney disease can be prevented and treated if known earlier. One way to find out chronic kidney failure is to detect using data mining. Iterative Dichotomiser 3 (ID3) algorithm is one of the classification methods and is a type of method that can map or separate two or more different classes. Based on the measurement of performance classification of 80% of training data from 400 data used, it shows that the accuracy value reached 96.25%. It can be concluded that the ID3 Algorithm method is feasible to be used in research predictions for chronic kidney failure.