Wyniki wyszukiwania - BazTech

1

Predicting banking stock prices using RNN, LSTM, and GRU approach

Satria Dias

Applied Computer Science

|

2023

|

Vol. 19, no 1

82--94

EN

In recent years, the implementation of machine learning applications started to apply in other possible fields, such as economics, especially investment. But, many methods and modeling are used without knowing the most suitable one for predicting particular data. This study aims to find the most suitable model for predicting stock prices using statistical learning with Arima Box-Jenkins, RNN, LSTM, and GRU deep learning methods using stock price data for 4 (four) major banks in Indonesia, namely BRI, BNI, BCA, and Mandiri, from 2013 to 2022. The result showed that the ARIMA Box-Jenkins modeling is unsuitable for predicting BRI, BNI, BCA, and Bank Mandiri stock prices. In comparison, GRU presented the best performance in the case of predicting the stock prices of BRI, BNI, BCA, and Bank Mandiri. The limitation of this research was data type was only time series data. It limits our instrument to four statistical methode only.

2

Performance evaluation of stock price prediction models using EMAGRU

Erizal Erizal, Diqi Mohammad

Applied Computer Science

|

2023

|

Vol. 19, no 3

160--173

EN

Stock price prediction is an exciting issue and is very much needed by investors and business people to develop their assets. The main difficulties in predicting stock prices are dynamic movements, high volatility, and noises caused by company performance and external influences. The traditional method investors use is the technical analysis based on statistics, valuation of previous stock portfolios, and news from the mass media and social media. Deep learning can predict stock price movements more accurately than traditional methods. As a solution to the issue of stock prediction, the authors offer the Exponential Moving Average Gated Recurrent Unit (EMAGRU) model and demonstrate its utility. The EMAGRU architecture contains two stacked GRUs arranged in parallel. The inputs and outputs are the EMA10 and EMA20, formed from the closing prices over ten years. The authors also combine the AntiReLU and ReLU activation functions into the model so that EMAGRU has 6 model variants. The proposed model produces low losses and high accuracy. RMSE, MEPA, MAE, and R^2 are 0.0060, 0.0064, 0.0050, and 0.9976 for EMA10, and 0.0050, 0.0058, 0.0045, and 0.9982 for EMA20, respectively.

3

Predicting hospital emergency department visits with deep learning approaches

Zhao Xinxing, Lai Joel Weijia, Ho Andrew Fu Wah, Liu Nan, Ong Marcus Eng Hock, Cheong Kang Hao

Biocybernetics and Biomedical Engineering

|

2022

|

Vol. 42, no. 3

1051--1065

EN

Overcrowding in emergency department (ED) causes lengthy waiting times, reduces adequate emergency care and increases rate of mortality. Accurate prediction of daily ED visits and allocating resources in advance is one of the solutions to ED overcrowding problem. In this paper, a deep stacked architecture is being proposed and applied to the daily ED visits prediction problem with deep components such as Long Short Term Memory (LSTM), Gated Recurrent Units (GRU) and simple Recurrent Neural Network (RNN). The proposed architecture achieves very high mean accuracy level (94.28–94.59%) in daily ED visits predictions. We have also compared the performance of this architecture with non-stacked deep models and traditional prediction models. The results indicate that deep stacked models outperform (4–7%) the traditional prediction models and other non-stacked deep learning models (1–2%) in our prediction tasks. The application of deep neural network in ED visits prediction is novel as this is one of the first studies to apply a deep stacked architecture in this field. Importantly, our models have achieved better prediction accuracy (in one case comparable) than the state-of-the-art in the literature.

4

The hydrological impact of tropical cyclones on soil moisture using a sensor based hybrid deep learning model

Srivastava Gautam, Kavitha V., Vimaladevi M., Parthiban N.

Acta Geophysica

|

2022

|

Vol. 70, no 6

2933--2951

EN

Tropical cyclones that originate from the Indian Ocean affect the Indian Sub-Continent. Heavy rainfall and flooding occur because of these cyclones. South Odisha was affected by Cyclonic Storm Daye in September 2018 and Cyclonic Storm Titli was occurred in August affecting Andhra Pradesh and Odisha as well. The Eastern portion of India was affected by the Cyclonic Storm Fani in April 2019. In May 2020, West Bengal was affected by the Amphan which is a Super Cyclonic Storm and in the same year Tamil Nadu was affected by the very severe Cyclonic Storm Nivar in November 2020. These are just a few of the notable cyclonic events in the Indian Sub-Continent. These cyclonic events cause a dramatic change in a very short time from dry soil to exceptional flooding. In this proposed work, we are attempting to create an observations-driven prediction model to quantify the soil moisture variations daily, predict county-based meteorology and evaluate the cause of cyclones and heavy rainfall in certain areas of India. In our work, we applied a deep learning-based methodology to predict soil moisture. For the prediction model, we fused Feed Forward Neural Networks with the Gated Recurrent Unit (GRU) model and present the prediction results. We have used climatic as well as environmental data published by the Indian Meteorological Department (IMD) Warning from 2011. The collected data is time-series data. Comparisons and the relationship that exists between soil moisture and meteorological data are made and analyzed. The soil moisture of the South Indian states Karnataka, Andhra Pradesh and Tamil Nadu are predicted from weather data using a hybrid deep learning model. The evaluations of the proposed work using Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and R-squared (R2 ) against Non-hybrid Neural Network models such as Artificial Neural Networks (ANN), Convolutional Neural Networks, and Gated Recurrent Unit (GRU) models is analyzes where our model has given better results.

5

An optimized parallel implementation of non-iteratively trained recurrent neural networks

El Zini Julia, Rizk Yara, Awad Mariette

Journal of Artificial Intelligence and Soft Computing Research

|

2021

|

Vol. 11, No. 1

33--50

EN

Recurrent neural networks (RNN) have been successfully applied to various sequential decision-making tasks, natural language processing applications, and time-series predictions. Such networks are usually trained through back-propagation through time (BPTT) which is prohibitively expensive, especially when the length of the time dependencies and the number of hidden neurons increase. To reduce the training time, extreme learning machines (ELMs) have been recently applied to RNN training, reaching a 99% speedup on some applications. Due to its non-iterative nature, ELM training, when parallelized, has the potential to reach higher speedups than BPTT. In this work, we present Opt-PR-ELM, an optimized parallel RNN training algorithm based on ELM that takes advantage of the GPU shared memory and of parallel QR factorization algorithms to efficiently reach optimal solutions. The theoretical analysis of the proposed algorithm is presented on six RNN architectures, including LSTM and GRU, and its performance is empirically tested on ten time-series prediction applications. Opt- PR-ELM is shown to reach up to 461 times speedup over its sequential counterpart and to require up to 20x less time to train than parallel BPTT. Such high speedups over new generation CPUs are extremely crucial in real-time applications and IoT environments.

6

Detection and Localization of Audio Event for Home Surveillance Using CRNN

Suruthhi V. S., Smita V., Rolant Gini J., Ramachandran K. I.

International Journal of Electronics and Telecommunications

|

2021

|

Vol. 67, No. 4

735--741

EN

Safety and security have been a prime priority in people’s lives, and having a surveillance system at home keeps people and their property more secured. In this paper, an audio surveillance system has been proposed that does both the detection and localization of the audio or sound events. The combined task of detecting and localizing the audio events is known as Sound Event Localization and Detection (SELD). The SELD in this work is executed through Convolutional Recurrent Neural Network (CRNN) architecture. CRNN is a stacked layer of convolutional neural network (CNN), recurrent neural network (RNN) and fully connected neural network (FNN). The CRNN takes multichannel audio as input, extracts features and does the detection and localization of the input audio events in parallel. The SELD results obtained by CRNN with the gated recurrent unit (GRU) and with long short-term memory (LSTM) unit are compared and discussed in this paper. The SELD results of CRNN with LSTM unit gives 75% F1 score and 82.8% frame recall for one overlapping sound. Therefore, the proposed audio surveillance system that uses LSTM unit produces better detection and overall performance for one overlapping sound.

7

Blood glucose prediction with deep neural networks using weighted decision level fusion

Dudukcu Hatice Vildan, Taskiran Murat, Yildirim Tulay

Biocybernetics and Biomedical Engineering

|

2021

|

Vol. 41, no. 3

1208--1223

EN

Background and Objective: Diabetes mellitus is a chronic disease that requires regular monitoring of blood glucose in the circulatory system. If the amount of glucose in the blood is not regulated constantly, this may have vital consequences for the individual. For this reason, there are many studies in the literature that perform blood glucose (BG) prediction. Methods: Blood glucose prediction is generally performed by using many parameters. In this paper, it was attempted to predict the future blood glucose values of the patient by using only the blood glucose values of diabetes patients’ history. For this purpose, Long short term memory (LSTM), WaveNet and Gated Recurrent Units (GRU) and decision-level combinations of these architectures were used to predict blood glucose. First of all, hyper-parameters were selected for the most efficient operation of these network architectures and experimental studies were conducted using the extended OhioT1DM data set which has blood glucose history of 12 diabetes patients. Results: Experimental studies using 30, 45 and 60 min prediction horizon (PH), the average lowest RMSE value were obtained by the fusion of three networks as 21.90 mg/dl, 29.12 mg/dl, 35.10 mg/dl respectively. Conclusions: When the obtained RMSE value compared to state-of-art studies in the literature, the results show that the proposed method is quite successful for short-term blood glucose prediction. In addition, the proposed fusion method gives a new perspective for future studies in the literature for BG prediction.

8

Forecasting future values of time series using the lstm network on the example of currencies and WIG20 companies

Mróz Bartosz, Nowicki Filip

Zeszyty Naukowe. Telekomunikacja i Elektronika / Uniwersytet Technologiczno-Przyrodniczy w Bydgoszczy

|

2020

|

nr 24

13--30

EN

The article presents a comparison of the RNN, GRU and LSTM networks in predicting future values of time series on the example of currencies and listed companies. The stages of creating an application which is a implementation of the analyzed issue were also shown – the selection of networks, technologies, selection of optimal network parameters. Additionally, two conducted experiments were discussed. The first was to predict the next values of WIG20 companies, exchange rates and cryptocurrencies. The second was based on investments in cryptocurrencies guided solely by the predictions of artificial intelligence. This was to check whether the investments guided by the predictions of such a program have a chance of effective earnings. The discussion of the results of the experiment includes an analysis of various interesting phenomena that occurred during its duration and a comprehensive presentation of the relatively high efficiency of the proposed solution, along with all kinds of graphs and comparisons with real data. The difficulties that occurred during the experiments, such as coronavirus or socio-economic events, such as riots in the USA, were also analyzed. Finally, elements were proposed that should be improved or included in future versions of the solution – taking into account world events, market anomalies and the use of supervised learning.

PL

W artykule przedstawiono porównanie sieci RNN, GRU i LSTM w przewidywaniu przyszłych wartości szeregów czasowych na przykładzie walut i spółek giełdowych. Przedstawiono również etapy tworzenia aplikacji będącej realizacją analizowanego zagadnienia – dobór sieci, technologii, dobór optymalnych parametrów sieci. Dodatkowo omówiono dwa przeprowadzone eksperymenty. Pierwszym było przewidywanie kolejnych wartości spółek z WIG20, kursów walut i kryptowalut. Drugi opierał się na inwestycjach w kryptowaluty, kierując się wyłącznie przewidywaniami sztucznej inteligencji. Miało to na celu sprawdzenie, czy inwestowanie na podstawie przewidywania takiego programu pozwala na efektywne zarobki. Omówienie wyników eksperymentu obejmuje analizę różnych ciekawych zjawisk, które wystąpiły w czasie jego trwania oraz kompleksowe przedstawienie relatywnie wysokiej skuteczności proponowanego rozwiązania wraz z wszelkiego rodzaju wykresami i porównaniami z rzeczywistymi danymi. Analizowano również trudności, które wystąpiły podczas eksperymentów, takie jak koronawirus, wydarzenia społeczno-gospodarcze czy zamieszki w USA. Na koniec zaproponowano elementy, które należałoby ulepszyć lub uwzględnić w przyszłych wersjach rozwiązania, uwzględniając wydarzenia na świecie, anomalie rynkowe oraz wykorzystanie uczenia się nadzorowanego.

9

Towards a new deep learning algorithm based on GRU and CNN : NGRU

Atassi Abdelhamid, el Azami Ikram

Journal of Automation Mobile Robotics and Intelligent Systems

|

2020

|

Vol. 14, No. 4

45--47

EN

This paper describes our new deep learning system based on a comparison between GRU and CNN. Initially we start with the first system which uses Convolutional Neural Network (CNN) which we will compare with the second system which uses Gated Recurrent Unit (GRU). And through this comparison we propose a new system based on the positive points of the two previous systems. Therefore, this new system will take the right choice of hyper-parameters recommended by the authors of both systems. At the final stage we propose a method to apply this new system to the dataset of different languages (used especially in socials networks).

10

Rosyjskie wpływy w Niemczech

Romanowska A.

Wiedza Obronna

|

2018

|

nr 1-2

32--57

PL

Aktualnie w rosyjskich działaniach mających na celu wpływ na politykę Zachodu pojawiły się trzy nowe aspekty. Po pierwsze, są one skierowane nie tylko na peryferia Europy czy na poszczególne narody, takie jak Niemcy czy Polska, lecz także na destabilizację (od wewnątrz) całego projektu Unii Europejskiej (UE). Po drugie, zarówno zamaskowane, jak i jawne „aktywne środki” działania rosyjskiego państwa, są znacznie bardziej różnorodne, realizowane na większą skalę i bardziej zaawansowane technologicznie; nieustannie dopasowują się do zmieniającej się sytuacji. Po trzecie, uderzając jednocześnie w Europę i Stany Zjednoczone (USA), ingerencja Rosji zdaje się być skierowana na podważenie podstawowych wartości Zachodu. Celem tej strategii jest zakwestionowanie spójności i skuteczności Zachodu jako normatywnej siły utrzymującej porządek globalny oparty na uniwersalnych zasadach, a nie na sile. Dla Rosji, wpływy w Niemczech są priorytetem: osłabiając ten kraj, Rosja może skutecznie rozbijać UE jako instytucję oraz jedność państw europejskich, a także starać się zerwać transatlantycką więź między Unią a Stanami Zjednoczonymi. Rosyjskie służby wywiadowcze są aktywne od wielu lat, z dużą intensywnością działają przeciwko niemieckim interesom zarówno w ich kraju, jak i w Rosji, i nie ma powodu, by zakładać, że ich działalność szpiegowska zmniejszy się w przewidywalnej przyszłości. Niemcy prowadzą negocjacje z Rosją i Ukrainą w ramach procesu mińskiego, a także organizują i podtrzymują europejski konsensus w sprawie sankcji wobec Rosji. Działania te sprawiają, że Berlin jest główną przeszkodą dla Rosji w realizacji jej interesów w Europie i na Ukrainie. Temat ten był dostrzegany przez niemieckie think tanki i media przez ostatnie cztery lata – zbiegł się z aneksją Krymu i początkiem wzmożonej rosyjskiej aktywności, realizowanej na dużą skalę w niemieckich mediach społecznościowych. Przedmiotem analiz stały się również różne aspekty organizacji życia mniejszości rosyjskiej w Niemczech.

EN

Three things are new about Russian interference today. Firstly, it appears to be directed not just at Europes periphery, or at specific European nations like Germany or Poland, but at destabilizing the European project from the inside out. Secondly, its covert and overt active measures are much more diverse, larger-scale, and more technologically sophisticated; they continually adapt and morph in accordance with changing technology and circumstances. Thirdly, by striking at Europe and the United States at the same time, the interference appears to be geared towards undermining the effectiveness and cohesion of the Western alliance as such and at the legitimacy of the West as a normative force upholding a global order based on universal rules rather than might alone. For Russia, which is clearly focused on the destabilization of Europe and the Transatlantic Alliance, the influence in Germany is a priority: by weakening Germany, Russia can effectively break the EU as an institution and unity of European states, and also try to break the transatlantic link between the EU and the United States. A divided Germany was Ground Zero for espionage, propaganda, and other kinds of influence operations throughout the Cold War; this did not end with the fall of the Berlin Wall. The Russian intelligence services have been active for many years with high intensity against German interests in Germany and in the Russian Federation there is no reason to assume that their espionage activities will abate in the foreseeable future. Germany is negotiating with Russia and Ukraine in the scope of the Minsk process, and is organizing and upholding the European consensus on sanctions against Russia. These actions make Berlin the main obstacle for Russia in realizing its interests in Europe and Ukraine. The topic has been prominent on the radar of German think tanks and media for the past four years roughly coinciding with the annexation of Crimea and the beginning of large-scale Russian trolling in German social media. The various aspects of life of the Russian minority in Germany have also been a subject to analysis.