Search

Your search keyword '"time series prediction"' showing total 4 results
Start Over Descriptor "time series prediction" Language turkish
4 results on '"time series prediction"'

1. Güneş ışınımı tahmini için CNN-LSTM modeli: performans analizi.

Author

Eşlik, Ardan Hüseyin, Şen, Ozan, and Serttaş, Fatih

Abstract

To fully utilize the potential of solar energy and effectively operate solar energy systems, it is vital to know solar radiation values. Modeling solar radiation data with high variability is a complex problem, and nonlinear methods are needed. This study proposes a hybrid model using CNN and LSTM architectures for solar radiation prediction. The performance and applicability of the proposed model are examined by comparing it with ARIMA statistical method and different machine learning methods such as Random Forest, Decision Tree and K-Nearest Neighbor. The study used hourly solar radiation values measured with a pyranometer on the Afyon Kocatepe University campus. As a result of the experimental studies, the CNNLSTM model is the most successful model in terms of RMSE, MAE, MAPE and r² evaluation criteria, achieving 16.92%, 17.25%, 17.24%, 22.89% more successful results than the Decision Trees model, which produced the most unsuccessful predictions, respectively. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

2. Time Series Prediction with Digital Twins in Public Transportation Systems

Author

Mehmet Ali Ertürk

Subjects
digital twin, intelligent transportation systems, iot, time series prediction, Industrial engineering. Management engineering, T55.4-60.8, Business, HF5001-6182
Abstract

Classical traffic and transportation control centers must be more robust with the rapid spread of electric, intelligent, autonomous, and software-defined vehicles. Existing traffic management strategies have significant drawbacks in public safety, predictive maintenance, tuning the core functionality of vehicles, and managing mobility. We can renovate this system with next-generation intelligent Digital Twin (DT) technologies. This research proposes a time-series prediction system through Digital Twins to manage the public transportation system with Facebook’s Prophet. This study presents a model framework to build a Digital Twin application in Intelligent Public Transportation Systems and uses a public data set to validate the model with Facebook’s Prophet library by forecasting metro line passenger flows. According to the results, the Mean Absolute Percentage Error (MAPE) is 0.017 for a 1-day horizon.

Published
2023
Full Text
View/download PDF

3. NARX Neural Networks Based Time Series Prediction for Amount of Airworthiness Time

Author

Tatli, Ali, Kahvecioğlu, Sinem, and Anadolu Üniversitesi, Sivil Havacılık Yüksekokulu

Subjects
Airworthiness, Artificial Neural Networks Performance, Narx Networks, Time Series Prediction
Abstract

National Conference on Electrical, Electronics and Biomedical Engineering (ELECO) -- DEC 01-03, 2016 -- Bursa, TURKEY, WOS: 000401519800026, Aviation, a serious sector, is required to recourse planning intensively. There is a need for foresight in order to eliminate the planless that can lead to high costs and prevent accidents that could cause to fatal consequences. Without doubt, considering inexperienced pilot candidates this situation comes into prominence for profit-oriented flight training organizations. This study contains extracting time series of airworthiness and forecasting amount of daily airworthiness time based on NARX (Nonlineer Auto Regressive Exogenous) neural network models., Uludag Univ, Muhendislik Fakultesi, Elektrik Elektronik Muhendisligi Bolumu, Istanbul Teknik Univ, Elektrik Elektronik Fakultesi, TMMOB Elektrik Muhendisleri Odasi Bursa Subesi

Published
2016

4. Kısıtlanmış Boltzmann makinesi ile zaman serilerinin tahmini

Author

Pekmezci, Mehmet, Akman, Ali, Maltepe Üniversitesi, Fen Bilimleri Enstitüsü, Pekmezci, Mehmet, and Bilgisayar Mühendisliği Ana Bilim Dalı

Subjects
İleri yönelimli yapay sinir ağları, Time series, Boltzmann Makinesi, Zaman serileri tahmini, Feed forward artificial neural networks, Time series prediction, Boltzmann Machine, Computer Engineering and Computer Science and Control, Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol
Abstract

Bu çalışmada bir yapay sinir ağı ( çeşidi olan Kısıtlanmış Boltzmann Makinesi ile zaman serileri tahmini yapılmıştır. Karşılaştırma amaçlı olarak da, standart bilinen İleri Yönelimli YSA ile de tahmin yapılmıştır ve karşılaştırma sonuçları bu çalışma içinde grafik olarak verilmiştir. Bu çalışma için yazılmış olan kodlara “ http://java crbm backprop timeseries sample.googlecode.com/svn/ trunk/trunk/” adresinden ulaşılabilmektedir. Programı bilgisayara indirip Test sınıfını çalıştırarak karşılaştırma sonuçlarına ulaşılabilmektedir. Bu kodlar yazılırken “jaRBM (Ben Chung ,2009 ))” adlı açık kaynak kodlu yazılım kütüphanesinden yararlanılmıştır Bu tez 2012 yılında yapılmıştır ve 73 sayfadan oluşmaktadır., In this work, a type of artificial neural network called “Restr i cted Boltzmann Machine” is used to predict time series. Another type of artifi c i al neural network called “Feed Forward Network” is also used to predict the time series to be able to compare prediction results. Prediction results and their comparison are given as both table and graphical representation. The software codes written within this work can be reached at http://java crbm backprop timeseries sample.googlecode.com/svn/trunk/ trunk/” It is sufficient to download the program and run the “Test” class in the p rogram to obtain the prediction results, graphics and the i r comparison tables. Some code snippets of the open source java project “jaRBM (Ben Chung ,2009 ))” is also used in the software codes. This thesis written in 2012 and consists of 73 pages.

Published
2012

Catalog

Books, media, physical & digital resources
See catalog results