STOCK PRICE PREDICTION AND SIMULATION USING GEOMETRIC BROWNIAN MOTION-KALMAN FILTER: A COMPARISON BETWEEN KALMAN FILTER ALGORITHMS

The importance of the food sector to Indonesia's economy makes it one of the most attractive sectors to consider in an investment portfolio. An optimal portfolio is the best choice for investors among various efficient portfolios, aiming to maximize returns while minimizing risk. Moreover, since investment is inherently associated with fluctuating stock prices, accurate forecasting is necessary to anticipate future stock movements. This study aims to accurately predict stock prices and construct an optimal portfolio consisting of five food sector stocks listed on the Indonesia Stock Exchange, namely DMND, ICBP, HOKI, INDF, and ULTJ. Stock price predictions are generated using the Holt-Winter method, which can identify seasonal patterns and trends from historical data. The predicted stock prices are then used to calculate returns, which serve as the basis for portfolio optimization using the Mean-Variance model. The results show that the Holt-Winter method successfully produces accurate stock price forecasts, with Mean Absolute Percentage Error (MAPE) values for all stocks below 10%. These forecasts are used to calculate returns in the portfolio optimization process. The optimal portfolio composition is determined with the following weight proportions: HOKI (4%), ICBP (18%), ULTJ (21%), DMND (26%), and INDF (30%). This portfolio yields an expected return of 0.0441% and a portfolio variance of 0.0063%, reflecting a balanced trade-off between potential return and risk.

The stock market price movement has become an interesting topic to research due to the large profit potential. Market traders use technical analysis to forecast stock prices by analyzing stock transaction data. Because technical analysis predictions are subjective, an automated system is needed to produce more objective and faster results. In the previous, next prices were predicted by combining price and transaction volume features with technical analysis indicators such as Stochastic, Moving Average, Moving Average Convergence Divergence, and Relative Strength Index values. In most cases, the addition of the indicator's value reduces the accuracy of the forecast. This is because the value of technical analysis indicators is not useful enough when they are used directly without transformation. The transformation of technical analysis indicators to a crossing value between the fast and slow signal on these indicators, known as the golden cross and death cross, has a greater impact on stock price movement. Therefore, it is recommended to use the price and the golden cross and death cross on technical analysis indicators as inputs to improve stock price predictions. Long Short Term Memory is used in the proposed method, which is a reliable method for predicting sequence data. The stock price forecasts are based on 1200 days of trading data from the Indonesia Stock Exchange for 10 LQ45 group shares. The test was conducted by calculating the Root Mean Squared Error and the Mean Absolute Percentage Error value of the predicted close price next day.

Prediction of the stock price has always been a challenging task due to irregular patterns of the market. Uncertainty has made researchers think of some new and robust predictive methods. Many studies are available in the literature, with many models to predict the stock price accurately. Statistical, machine learning, deep learning, and other related approaches can create a predictive model. ARIMA model is the most commonly used statistical model for time series prediction. But ensemble learning techniques have not been explored much to predict future stock price. So, the present study stresses comparing statistical methods with ensemble learning methods. This paper compares the ARIMA, Random Forest, and Extreme Gradient Boosting models based on root mean squared error (RMSE) and mean absolute percentage error (MAPE). The subject chosen is Google’s stocks, and the data used is from NASDAQ stock exchange. The analysis results show that the ARIMA model performed fairly well for short-term predictions but relatively high MAPE value. The extreme gradient boosting model gave the best performance with the lowest RMSE and MAPE value. Hence, it is evident that after proper hyperparameter tuning, ensemble learning techniques can be used to create robust stock price-prediction models.

Sharia stocks are one of the investment instruments in the Islamic capital market. In the capital market, it is known that stock prices are very volatile. This makes investors need to carry out a strategy for making the right decision in investing, one of which can be done by predicting stock prices. In this study, predictions were made using historical data on the closing price of Islamic shares of PT. Telkom Indonesia Tbk with the Bidirectional Long Short Term Memory (BiLSTM) method. In building the best prediction model, it is necessary to choose the right parameters and one way to do this is to use the grid search algorithm. Based on the results of the test analysis, it was found that the smallest Mean Absolute Percentage Error (MAPE) value was found in the BiLSTM model in the distribution of data with a percentage of 90% training data and 10% testing data and parameter values obtained based on parameter tuning using grid search, including the number of neurons 25, 100 epochs, 4 batches, and 0.2 dropouts. The MAPE obtained in this study was 10.83% and based on the scale on the MAPE value criteria, this shows that the resulting prediction model is accurate. As for the test results from the comparisons made on the BiLSTM and LSTM models using grid search as a tuning parameter and models without using a grid search or it can be called a trial and error approach as a tuning parameter, it is found that the model with better predictive performance is found in BiLSTM using a grid search. compared to other models.

Seven-day test result assessment of the developed strength in composite cement mortar with slag

This study aims to design, implement, and compare the performance of the Backpropagation algorithm from Artificial Neural Networks and the Seasonal Autoregressive Integrated Moving Average (SARIMA) model in predicting the optimal daily rice requirement at Grillme Restaurant in Pontianak. The main problem faced by the restaurant is the uncertainty in determining the required daily rice stock, which periodically results in either understocking (shortage) or overstocking (wastage), leading to operational losses. To address this, the study utilizes historical daily rice sales data from January 2023 to April 2025 as the database for training and testing both predictive models. The SARIMA approach is employed to capture time series components (trend and seasonality), while Backpropagation is utilized to model non-linear patterns. Comparative test results indicate that the SARIMA model achieved superior accuracy compared to the Backpropagation model. This is confirmed by the Mean Absolute Percentage Error (MAPE) value of the SARIMA algorithm being 17.35%, which is lower than the MAPE value of Backpropagation at 19.62%. The MAPE values obtained by both models demonstrate good predictive capability, but it is concluded that SARIMA is more recommended for a more efficient and planned management of rice stock at Grillme Restaurant in Pontianak.

The demand for beef supplies depended on seasonal patterns because it depends on feed supplies, especially in rural areas that still rely on natural feed. Beef supplies were part of government regulations because they were highly demanded commodities. They are livestock products that contain nutritional value to meet the protein needs of the community. Beef stocks were influenced by factors such as beef production, beef consumption, and people's income levels. In anticipating the increasing demand for beef, it is necessary to carry out forecasting to estimate the demand for meat in the future. In forecasting, various methods were used to choose the method with the lowest error rate. This research will compare the Double Exponential Smoothing (DES) with the Double Moving Average (DMA) based on the Mean Absolute Percentage Error (MAPE) measurement, one method for finding the minor error value. Based on the test results with beef supplies in Madura and conducting analysis, it can be concluded that the method with the smallest MAPE value is the Double Exponential Smoothing method, with the smallest MAPE value of 9.50% at an alpha parameter of 0.5. In testing with the Double Moving Average method, by determining the best MAPE value, the best time order in the Double Moving Average method is at time order parameter 2 with a MAPE value of 29.8408%. After finding the parameter with the smallest MAPE value, that parameter is used for data testing. In the measurement, data testing for data of 1 year, two years, three years, and four years. Each method has a level of error value that increases the same; a lot of data entered can affect the size of the MAPE value. So, the more data entered, the smaller the resulting error value.

Financial market prediction under deep learning framework using auto encoder and kernel extreme learning machine

BackgroundWrist-worn devices that incorporate photoplethysmography (PPG) sensing represent an exciting means of measuring heart rate (HR). A number of studies have evaluated the accuracy of HR measurements produced by these devices in controlled laboratory environments. However, it is also important to establish the accuracy of measurements produced by these devices outside the laboratory, in real-world, consumer use conditions.ObjectiveThis study sought to examine the accuracy of HR measurements produced by the Withings ScanWatch during free-living activities.MethodsA sample of convenience of 7 participants volunteered (3 male and 4 female; mean age 64, SD 10 years; mean height 164, SD 4 cm; mean weight 77, SD 16 kg) to take part in this real-world validation study. Participants were instructed to wear the ScanWatch for a 12-hour period on their nondominant wrist as they went about their day-to-day activities. A Polar H10 heart rate sensor was used as the criterion measure of HR. Participants used a study diary to document activities undertaken during the 12-hour study period. These activities were classified according to the 11 following domains: desk work, eat or drink, exercise, gardening, household activities, self-care, shopping, sitting, sleep, travel, and walking. Validity was assessed using the Bland-Altman analysis, concordance correlation coefficient (CCC), and mean absolute percentage error (MAPE).ResultsAcross all activity domains, the ScanWatch measured HR with MAPE values <10%, except for the shopping activity domain (MAPE=10.8%). The activity domains that were more sedentary in nature (eg, desk work, eat or drink, and sitting) produced the most accurate HR measurements with a small mean bias and MAPE values <5%. Moderate to strong correlations (CCC=0.526-0.783) were observed between devices for all activity domains, except during the walking activity domain, which demonstrated a weak correlation (CCC=0.164) between devices.ConclusionsThe results of this study show that the ScanWatch measures HR with a degree of accuracy that is acceptable for general consumer use; however, it would not be suitable in circumstances where more accurate measurements of HR are required, such as in health care or in clinical trials. Overall, the ScanWatch was less accurate at measuring HR during ambulatory activities (eg, walking, gardening, and household activities) compared to more sedentary activities (eg, desk work, eat or drink, and sitting). Further larger-scale studies examining this device in different populations and during different activities are required.

This paper proposes a way to forecast the future closing price of small size companies in Bursa Malaysia by using geometric Brownian motion (GBM). Forecasting is restricted to short term investment because most of the investors aim to gain profit in short period of time. The reasons of choosing small size companies are because the asset prices are lower, hence the asset are affordable for all level of investors. In this paper, we suggest that GBM which involves randomness, volatility, and drift can help investor in making their investment decision wisely. This research shows the model is highly accurate model in forecasting stock prices and it is proven by the lower value of mean absolute percentage error (MAPE). Although it is highly accurate, we try to find the suitable volatility measurements that give the forecast value closer to the actual movement of stock prices. The result shows that by using highs-lows-close volatility, the forecast stock prices are closest to the actual prices. This volatility measurement and GBM model are suggested to the investor to forecast future prices for a maximum of two week investment.

Fisheries is one of the sources of food security in Indonesia which plays an important role in the rich biodiversity of Indonesian coastal waters. This research uses the Chen Fuzzy Time Series Model forecasting method. The samples used in this study are data on the number of fishery commodity products with specific Tilapia and Catfish monthly Aceh Province as much as 24 data, starting from January 2021 to December 2022. The results of forecasting the number of tilapia in January 2023 were 1,320,815, with a MAPE (Mean Absolute Percentage Error) value of 12.93%. And the result of forecasting the number of catfish in January 2023 was 2,339,568, with a MAPE (Mean Absolute Percentage Error) value of 22.9%. This shows that forecasting using the Chen model Fuzzy Time Series method is good enough to forecast the number of tilapia and catfish in Aceh Province.

Modeling and forecasting stock prices of public corporates are important studies in financial analysis, due to their stock price characteristics. Stocks investments give a wide variety of risks. Taking a portfolio of several stocks is one way to minimize risk. Stochastic process of single stock price movements model can be formulated in Geometric Brownian Motion (GBM) model. But for a portfolio that consist more than one corporate stock, we need an expansion of GBM Model. In this paper, we use multidimensional Geometric Brownian Motion model. This paper aims to model and forecast two stock prices in a portfolio. These are PT. Matahari Department Store Tbk and PT. Telekomunikasi Indonesia Tbk on period January 4, 2016 until April 21, 2017. The goodness of stock price forecast value is based on Mean Absolute Percentage Error (MAPE). As the results, we conclude that forecast two stock prices in a portfolio using multidimensional GBM give less MAPE than using GBM for single stock price respectively. We conclude that multidimensional GBM is more appropriate for modeling stock prices, because the price of each stock affects each other.

Application of the Internet of things (IoT) for data collection in solar drying can be very efficient in collecting big data of drying parameters. There are many variables involved so it is hard to find a model to predict the moisture content of the food product during drying. In model building, interaction terms should be incorporated because they also contribute to the model. Eight selection criteria (8SC) is a very useful method in model building. This study applied ordinary least squares (OLS) regression and ridge regression with 8SC in model building to predict the moisture content of drying fish. A total of eighty models were considered in this study. One best model was chosen each from OLS regression and ridge regression. M78.7.3 with a total of eleven independent variables was the best OLS model after conducting multicollinearity and coefficient test. Next, the best ridge model M56.0.0 was obtained after the coefficient test. The mean absolute percentage error (MAPE) was used to measure the accuracy of the prediction model. For OLS model M78.7.3, the MAPE value was 15.7342. The MAPE value for ridge model M56.0.0 was 17.4054. From the MAPE value, OLS model M78.7.3 provided a better estimation than the ridge model M56.0.0. However, OLS model M78.7.3 violated the normality assumptions of residuals. This is highly caused by the outlier problem. So, due to non- normality of the residuals and presence of outliers in the dataset, ridge regression is preferred for the best forecast model.

The exchange rate between two countries is the price of the currency used by residents of these countries to trade with each other, the relationship between the Rupiah exchange rate and the Yuan is one of the important aspects in the dynamics of international trade. Therefore, forecasting the exchange rate is important as an effort to predict the exchange rate of Rupiah against Yuan in the future. The method used for forecasting is Singular Spectrum Analysis, namely decomposition and reconstruction. The accuracy of the resulting forecast is measured using the Mean Absolute Percentage Error criterion. The exploration results obtained are forecasting accuracy based on the Mean Absolute Percentage Error value of 2.15% with a window length of 23 which identifies that the forecasting results are accurate and effective. Forecasting is said to be accurate if the Mean Absolute Percentage Error value is lower than 10% and close to 10%

This research uses Facebook's Prophet algorithm to create a web-based stock price prediction system for the Indonesia Stock Exchange's (IDX) LQ45 index. From an economic standpoint, this study explores how easily available predictive analytics might impact market dynamics in emerging economies and empower individual investors. The program, which was created with Streamlit and Python, analyzes historical data going back to 2018. The system exhibits strong predicting capabilities and great user acceptance, even though specific average Mean Absolute Percentage Error (MAPE) values need to be further characterized. The application's exceptional usability and potential for broad adoption are highlighted by the User Acceptance Test (UAT) results, which show an impressive 94.7% favorable response from 25 respondents. According to this research, democratizing advanced forecasting techniques might boost investor confidence, drastically lessen information asymmetry, and even cast doubt on the weak-form Efficient Market Hypothesis (EMH) in behaviorally biased markets. This study provides useful information for policymakers seeking to promote equitable capital market development and better economic decision-making by integrating technical innovation with financial economics.