Auflistung nach Schlagwort "Forecasting"
1 - 7 von 7
Treffer pro Seite
Sortieroptionen
- ZeitschriftenartikelA Simulation-Based Approach to Understanding the Wisdom of Crowds Phenomenon in Aggregating Expert Judgment(Business & Information Systems Engineering: Vol. 63, No. 4, 2021) Afflerbach, Patrick; Dun, Christopher; Gimpel, Henner; Parak, Dominik; Seyfried, JohannesResearch has shown that aggregation of independent expert judgments significantly improves the quality of forecasts as compared to individual expert forecasts. This “wisdom of crowds?? (WOC) has sparked substantial interest. However, previous studies on strengths and weaknesses of aggregation algorithms have been restricted by limited empirical data and analytical complexity. Based on a comprehensive analysis of existing knowledge on WOC and aggregation algorithms, this paper describes the design and implementation of a static stochastic simulation model to emulate WOC scenarios with a wide range of parameters. The model has been thoroughly evaluated: the assumptions are validated against propositions derived from literature, and the model has a computational representation. The applicability of the model is demonstrated by investigating aggregation algorithm behavior on a detailed level, by assessing aggregation algorithm performance, and by exploring previously undiscovered suppositions on WOC. The simulation model helps expand the understanding of WOC, where previous research was restricted. Additionally, it gives directions for developing aggregation algorithms and contributes to a general understanding of the WOC phenomenon.
- TextdokumentForecasting BEV charging station occupancy at work places(INFORMATIK 2020, 2021) Motz, Marvin; Huber, Julian; Weinhardt, ChristofAt many charging stations, the charging process of battery electric vehicles (BEV) takes significantly more time than refilling a gas tank. In combination with the lack of charging stations, this results in more planning effort for drivers who have to find a free charging station. In addition, charging draws a significant amount of energy from the power grid so that operators might have to coordinate charging to avoid congestion. Such problems are especially relevant at workplaces, where the employer might offer many charging stations for employees with similar working hours. An approach to overcome these problems lies in the management of the existing infrastructure using data-driven strategies. Accurate forecasts on the occupancy of charging stations allow allocating available resources more efficiently. This work aims to find suitable methods to predict the occupancy of single charging stations, given their historical data. The forecasts could be used as an input in decision support for drivers or energy management systems of charging station operators. This paper discusses feature importance, transferability between multiple charging stations at one location, and how the characteristics of charging stations influence the predictability of their occupancy. We use 52 charging stations from the open ACN data set to evaluate the research questions. The data set has more than 24,000 charging events and is located at a research facility so that it resembles workplace parking. Finally, we test the forecast on new, previously unseen data to ensure the findings hold-up in a realistic scenario.
- KonferenzbeitragA framework for capturing, statistically modeling and analyzing the evolution of software models(Software Engineering und Software Management 2018, 2018) Shariat Yazdi, Hamed; Angelis, Lefteris; Kehrer, Timo; Kelter, UdoIn this work, we report about a recently developed framework for capturing, statistically modeling and analyzing the evolution of software models, published in the Journal of Systems and Software, Vol-118, Aug-2016. State-of-the-art approaches to understand the evolution of models of software systems are based on software metrics and similar static properties; the extent of the changes between revisions of a software system is expressed as differences of metrics values, and statistical analyses are based on these differences. Unfortunately, such approaches do not properly reflect the dynamic nature of changes. In contrast to this, our framework captures the changes between revisions of models in terms of both low-level (internal) and high-level (developer-visible) edit operations applied between revisions. Evolution is modeled statistically by using ARMA, GARCH and mixed ARMA-GARCH time series models. Forecasting and simulation aspects of these time series models are thoroughly assessed, and the suitability of the framework is shown by applying it to a large set of design models of real Java systems. A main motivation for, and application of, the resulting statistical models is to control the generation of realistic model histories which are intended to be used for testing model versioning tools. Further usages of the statistical models include various forecasting and simulation tasks.
- ZeitschriftenartikelRecurrent Neural Networks for Industrial Procurement Decisions(KI - Künstliche Intelligenz: Vol. 26, No. 4, 2012) Zimmermann, Hans-Georg; Tietz, Christoph; Grothmann, Ralph; Runkler, ThomasRational decisions are based upon forecasts. Precise forecasting has therefore a central role in business. The prediction of commodity prices or the prediction of energy load curves are prime examples. We introduce recurrent neural networks to model economic or industrial dynamic systems.
- ZeitschriftenartikelReducing energy time series for energy system models via self-organizing maps(it - Information Technology: Vol. 61, No. 2-3, 2019) Yilmaz, Hasan Ümitcan; Fouché, Edouard; Dengiz, Thomas; Krauß, Lucas; Keles, Dogan; Fichtner, WolfThe recent development of renewable energy sources (RES) challenges energy systems and opens many new research questions. Energy System Models (ESM) are important tools to study these problems. However, including RES into ESM strongly increases the model complexity, because one needs to model the fluctuant, weather-dependent electricity production from RES with a high level of granularity. This leads to long execution times. To deal with this issue, our objective is to reduce the input time series of ESM without losing their energy-related key characteristics, such as weather-dependent fluctuations in production or peak demands. This task is challenging, because of the variety and high-dimensionality of the data. We describe a carefully engineered data-processing pipeline to reduce energy time series. We use Self-Organizing Maps, a specific kind of neural network, to select “representative days”. We show that our approach outperforms the existing ones with respect to the quality of ESM results, and leads to a significant reduction of ESM execution times.
- ZeitschriftenartikelUsing Twitter to Predict the Stock Market(Business & Information Systems Engineering: Vol. 57, No. 4, 2015) Nofer, Michael; Hinz, OliverBehavioral finance researchers have shown that the stock market can be driven by emotions of market participants. In a number of recent studies mood levels have been extracted from Social Media applications in order to predict stock returns. The paper tries to replicate these findings by measuring the mood states on Twitter. The sample consists of roughly 100 million tweets that were published in Germany between January, 2011 and November, 2013. In a first analysis, a significant relationship between aggregate Twitter mood states and the stock market is not found. However, further analyses also consider mood contagion by integrating the number of Twitter followers into the analysis. The results show that it is necessary to take into account the spread of mood states among Internet users. Based on the results in the training period, a trading strategy for the German stock market is created. The portfolio increases by up to 36 % within a six-month period after the consideration of transaction costs.
- KonferenzbeitragVisualization and Machine Learning for Data Center Management(INFORMATIK 2019: 50 Jahre Gesellschaft für Informatik – Informatik für Gesellschaft (Workshop-Beiträge), 2019) Chircu, Alina; Sultanow, Eldar; Baum, David; Koch, Christian; Seßler, MatthiasIn this paper, we present a novel tool for data center management that incorporates data visualization and machine learning capabilities. We developed the tool in the context of an action design research project conducted at a large government agency in Germany, which hosts three highly available data centers containing more than 10,000 servers. We derived the requirements for the tool from qualitative interviews with agency employees who are familiar with monitoring the data center infrastructure as well as from a review of existing data center and other large infrastructure monitoring solutions. We implemented a web-based 3D prototype for the tool as an Angular 6 application running on Node.js, and evaluated it with the same employees. Most participants preferred the new tool, which provided a significantly better option and enabled visualization of historical data for all server instances at the same time, as well as real-time charts. Planned improvements will take advantage of the full potential of machine learning for time series forecasting.