1. Models for forecasting the traffic flow within the city of LjubljanaGašper Petelin, Rok Hribar, Gregor Papa, 2023, izvirni znanstveni članek Povzetek: Efficient traffic management is essential in modern urban areas. The development of intelligent traffic flow prediction systems can help to reduce travel times and maximize road capacity utilization. However, accurately modeling complex spatiotemporal dependencies can be a difficult task, especially when real-time data collection is not possible. This study aims to tackle this challenge by proposing a solution that incorporates extensive feature engineering to combine historical traffic patterns with covariates such as weather data and public holidays. The proposed approach is assessed using a new real-world data set of traffic patterns collected in Ljubljana, Slovenia. The constructed models are evaluated for their accuracy and hyperparameter sensitivity, providing insights into their performance. By providing practical solutions for real-world scenarios, the proposed approach offers an effective means to improve traffic flow prediction without relying on real-time data.
Ključne besede: traffic modeling, time-series forecasting, traffic-count data set, machine learning, model comparison
Objavljeno v DiRROS: 28.09.2023; Ogledov: 165; Prenosov: 65
 Celotno besedilo (5,05 MB)
Gradivo ima več datotek! Več... |
2. MsGEN : measuring generalization of nutrient value prediction across different recipe datasetsGordana Ispirova, Tome Eftimov, Sašo Džeroski, Barbara Koroušić-Seljak, 2023, izvirni znanstveni članek Povzetek: In this study, we estimate the generalization of the performance of previously proposed predictive models for nutrient value prediction across different recipe datasets. For this purpose, we introduce a quantitative indicator that determines the level of generalization of using the developed predictive model for new unseen data not presented in the training process. On a predefined corpus of recipe embeddings from six publicly available recipe datasets (i.e., projecting them in the same meta-feature vector space), we train predictive models on one of the six recipe datasets and test the models on the rest of the datasets. In parallel, we define and calculate generalizability indexes which are numbers that indicate how generalizable a predictive model is i.e., how well will a predictive model learned on one dataset perform on another one not involved in the training. The evaluation results prove the validity of these indexes – their relation with the accuracy of the predictions. Further, we define three sampling techniques for selecting representative data instances that will cover all parts from the feature space uniformly (involving data from all datasets) and further will improve the generalization of a predictive model. We train predictive models with these generalized datasets and test them on instances from the six recipe datasets that are not selected and included in the generalized datasets. The results from the evaluation of these predictive models show improvement compared to the results from the predictive models trained on one recipe dataset and tested on the others separately.
Ključne besede: ML pipeline, predictive modeling, nutrient prediction, recipe datasets
Objavljeno v DiRROS: 25.09.2023; Ogledov: 210; Prenosov: 111
Celotno besedilo (3,27 MB)
Gradivo ima več datotek! Več... |
3. Effects of specific parameters on simulations of energy use and air temperatures in offices equipped with radiant heating/cooling panelsSabina Jordan, Jože Hafner, Martina Zbašnik-Senegačnik, Andraž Legat, 2019, izvirni znanstveni članek Povzetek: When creating a simulation model to assess the performance of buildings, there is usually a lack of feedback information. Only in the case of measurements of a real building is a direct comparison of the measured values and simulated results possible. Parameter data related to users’ behavior or other events can also be obtained. Their evaluated frequency, magnitude and duration, along with boundary conditions, are crucial for the results. It is clear that none of them can be predicted very accurately. Most of them, however, are needed for computer modeling. In this paper we analyzed the well-defined TRNSYS simulation model of offices equipped with radiant ceiling panels for heating and cooling. The model was based on real case offices and was validated based on measurements for 1 year. The analysis included simulations in order to define what effect the parameters related mainly to users have on the energy use and the indoor air temperatures. The study confirmed that specific human activities influence the annual energy use to a relatively small degree and that their effects often counteract. It also confirmed the even more important fact that although small, these activities can influence the thermal comfort of users. It is believed that despite the fact that this research was based on an analysis of offices equipped with radiant ceiling panels, most of the results could be applied generally.
Ključne besede: measurements, modeling, simulation, validation, analysis, energy use, temperature
Objavljeno v DiRROS: 15.09.2023; Ogledov: 122; Prenosov: 65
Celotno besedilo (1,83 MB)
Gradivo ima več datotek! Več... |
4. Effect of surface oxidation on oxidative propane dehydrogenation over chromia : ǂan ǂab initio multiscale kinetic studyMatej Huš, Drejc Kopač, David Bajec, Blaž Likozar, 2021, izvirni znanstveni članek Ključne besede: dehydrogenation, chromium oxide, surface oxidation, propane, multiscale modeling
Objavljeno v DiRROS: 03.09.2021; Ogledov: 901; Prenosov: 710
Celotno besedilo (4,46 MB)
Gradivo ima več datotek! Več... |
5. Modeling of time consumption for selective and situational precommercial thinning in mountain beech forest standsDomen Arnič, Janez Krč, Jurij Diaci, 2021, izvirni znanstveni članek Povzetek: Rationalization and optimization of work is becoming increasingly important in the European forestry sector. In this study a tool for modeling three different precommercial thinning approaches in young beech mountain stands was developed based on several field studies. The simulation examines three primary types of precommercial thinning: selective thinning and two types of situational thinning. We studied the impact of the number of candidates/crop trees and the impact of harvesting intensity on the structure and consumption of productive time. We found that in terms of costs situational precommercial thinning is more rational than selective precommercial thinning, that harvesting intensity has a significant impact on time consumption and that the number of candidates or crop trees has a significant impact on time consumption as well as on the relationships between main and auxiliary productive time. The modeling has shown that situational thinning is an alternative to selective thinning and that, in addition to requiring smaller and more efficient harvesting machines, it offers a cost-effective and ergonomic option (more walking, less chainsaw operation) for the pre-commercial thinning of young forest.
Ključne besede: precommercial thinning, selective thinning, situational thinning, modeling, crop tree
Objavljeno v DiRROS: 22.03.2021; Ogledov: 1063; Prenosov: 626
Celotno besedilo (657,80 KB)
Gradivo ima več datotek! Več... |
6. A comparison of models for forecasting the residential natural gas demand of an urban areaRok Hribar, Primož Potočnik, Jurij Šilc, Gregor Papa, 2019, izvirni znanstveni članek Povzetek: Forecasting the residential natural gas demand for large groups of buildings is extremely important for efficient logistics in the energy sector. In this paper different forecast models for residential natural gas demand of an urban area were implemented and compared. The models forecast gas demand with hourly resolution up to 60 h into the future. The model forecasts are based on past temperatures, forecasted temperatures and time variables, which include markers for holidays and other occasional events. The models were trained and tested on gas-consumption data gathered in the city of Ljubljana, Slovenia. Machine-learning models were considered, such as linear regression, kernel machine and artificial neural network. Additionally, empirical models were developed based on data analysis. Two most accurate models were found to be recurrent neural network and linear regression model. In realistic setting such trained models can be used in conjunction with a weather-forecasting service to generate forecasts for future gas demand.
Ključne besede: demand forecasting, buildings, energy modeling, forecast accuracy, machine learning
Objavljeno v DiRROS: 15.03.2019; Ogledov: 2087; Prenosov: 998
Celotno besedilo (968,06 KB) |
