https://sdq.kastel.kit.edu/api.php?action=feedcontributions&user=If4991&feedformat=atomSDQ-Institutsseminar - Benutzerbeiträge [de]2024-03-29T06:29:08ZBenutzerbeiträgeMediaWiki 1.39.6https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Attention_Based_Selection_of_Log_Templates_for_Automatic_Log_Analysis&diff=2579Attention Based Selection of Log Templates for Automatic Log Analysis2023-07-18T20:16:03Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Vincenzo Pace<br />
|email=vincenzo.pace@mailbox.org<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2023-07-21<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=Log analysis serves as a crucial preprocessing step in text log data analysis, including anomaly detection in cloud system monitoring. However, selecting an optimal log parsing algorithm tailored to a specific task remains problematic.<br />
<br />
With many algorithms to choose from, each requiring proper parameterization, making an informed decision becomes difficult. Moreover, the selected algorithm is typically applied uniformly across the entire dataset, regardless of the specific data analysis task, often leading to suboptimal results.<br />
<br />
In this thesis, we evaluate a novel attention-based method for automating the selection of log parsing algorithms, aiming to improve data analysis outcomes. We build on the success of a recent Master Thesis, which introduced this attention-based method and demonstrated its promising results for a specific log parsing algorithm and dataset. The primary objective of our work is to evaluate the effectiveness of this approach across different algorithms and datasets.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2023-07-21&diff=2578Institutsseminar/2023-07-212023-07-18T20:12:24Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2023-07-21T11:30:00.000Z |raum=Raum 348 (Gebäude 50.34) }}“</p>
<hr />
<div>{{Termin<br />
|datum=2023-07-21T11:30:00.000Z<br />
|raum=Raum 348 (Gebäude 50.34)<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Generating_Causal_Domain_Knowledge_for_Cloud_Systems_Monitoring&diff=2469Generating Causal Domain Knowledge for Cloud Systems Monitoring2023-03-14T16:03:10Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Rakan Al Masri<br />
|email=rakanalmasri97@gmail.com<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2023-03-17-IPD-Boehm<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=While standard machine learning approaches rely solely on data to learn relevant patterns, in certain fields, this may not be sufficient. Researchers in the Healthcare domain, have successfully applied causal domain knowledge to improve prediction quality of machine learning models, especially for rare diseases. The causal domain knowledge informs the machine learning model about similar diseases, thus improving the quality of the predictions.<br />
<br />
However, some domains, such as Cloud Systems Monitoring, lack readily<br />
available causal domain knowledge, and thus the knowledge must be approximated.<br />
Therefore, it is important to have a systematic investigation of the processes and<br />
design decision that affect the knowledge generation process.<br />
<br />
In this study, we showed how causal discovery algorithms can be employed to generate causal domain knowledge<br />
from raw textual logs in the Cloud Systems Monitoring domain. We also<br />
investigated the impact of various design choices on the domain knowledge<br />
generation process through systematic testing across multiple datasets and<br />
shared the insights we gained. To our knowledge, this is the first time such an<br />
investigation has been conducted.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2023-03-17-IPD-Boehm&diff=2468Institutsseminar/2023-03-17-IPD-Boehm2023-03-14T15:57:39Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2023-03-17T11:30:00.000Z |raum=Raum 348 (Gebäude 50.34) }}“</p>
<hr />
<div>{{Termin<br />
|datum=2023-03-17T11:30:00.000Z<br />
|raum=Raum 348 (Gebäude 50.34)<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2023-03-24&diff=2467Institutsseminar/2023-03-242023-03-14T15:55:00Z<p>If4991: </p>
<hr />
<div>{{Termin<br />
|datum=2023-03-24T11:30:00.000Z<br />
|raum=Raum 348 (Gebäude 50.34)<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Generating_Causal_Domain_Knowledge_for_Cloud_Systems_Monitoring&diff=2362Generating Causal Domain Knowledge for Cloud Systems Monitoring2022-11-29T17:56:47Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Rakan Al Masri<br />
|email=rakanalmasri97@gmail.com<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-12-02 Zusatztermin<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=Recently the authors of the DomainML framework for Domain Knowledge Guided Machine Learning showed that heuristically generated hierarchical and textual domain knowledge could improve the performance of machine learning tasks in cloud system monitoring. <br />
<br />
However, they were unsuccessful in generating useful causal knowledge. The reason might be that the causal knowledge was generated with very simple heuristics. <br />
<br />
In this work, we plan to use causal learning algorithms to generate various forms of causal knowledge. We will then evaluate them on cloud system monitoring machine learning tasks with the DomainML framework.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Attention_Based_Selection_of_Log_Templates_for_Automatic_Log_Analysis&diff=2361Attention Based Selection of Log Templates for Automatic Log Analysis2022-11-29T17:44:27Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Vincenzo Pace |email=vincenzo.pace@mailbox.org |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2022-12-02 Zusatz…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Vincenzo Pace<br />
|email=vincenzo.pace@mailbox.org<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-12-02 Zusatztermin<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=Log parsing is an essential preprocessing step in text log data analysis, such as anomaly detection in cloud system monitoring.<br />
<br />
However, selecting the optimal log parsing algorithm for a concrete task is difficult. Many algorithms exist to choose from, and each needs proper parametrization. The selected algorithm applies to the whole dataset and with parameters determined independently of the data analysis task, which is not optimal.<br />
<br />
In this work, we evaluate a new attention-based method to automatically select the optimal log parsing algorithm for the data analysis task. The method was initially proposed in the Master Thesis and showed promising results on one log parsing algorithm and one dataset. In this work, we plan to test whether the algorithm is helpful for other algorithms and datasets.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Generating_Causal_Domain_Knowledge_for_Cloud_Systems_Monitoring&diff=2360Generating Causal Domain Knowledge for Cloud Systems Monitoring2022-11-29T17:08:30Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Rakan Al Masri |email=rakanalmasri97@gmail.com |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2022-12-02 Zusatz…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Rakan Al Masri<br />
|email=rakanalmasri97@gmail.com<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-12-02 Zusatztermin<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=Recently, researchers have shown that domain knowledge improves the performance of machine learning tasks. For example, in healthcare, using hierarchical taxonomies of symptoms improved the performance of risk prediction tasks, especially for rare diseases. Similar ideas proved to work also in other contexts, such as cloud system monitoring.<br />
<br />
The authors of the DomainML framework for Domain Knowledge Guided Machine Learning showed that generated hierarchical and textual domain knowledge could improve the performance of machine learning tasks in cloud system monitoring. However, they were unsuccessful in generating useful causal knowledge. The reason might be that the causal knowledge was generated with simple heuristics rather than actual causal learning algorithms.<br />
<br />
This thesis aims to generate various forms of causal knowledge and evaluate them on cloud system monitoring machine learning tasks within the DomainML framework for Domain Knowledge Guided Machine Learning.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2022-12-02_Zusatztermin&diff=2359Institutsseminar/2022-12-02 Zusatztermin2022-11-29T17:03:15Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2022-12-02T11:30:00.000Z |raum=Raum 010 (Gebäude 50.34) |online=https://kit-lecture.zoom.us/j/67744231815 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2022-12-02T11:30:00.000Z<br />
|raum=Raum 010 (Gebäude 50.34)<br />
|online=https://kit-lecture.zoom.us/j/67744231815<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Refining_Domain_Knowledge_for_Domain_Knowledge_Guided_Machine_Learning&diff=2292Refining Domain Knowledge for Domain Knowledge Guided Machine Learning2022-08-16T14:18:18Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Sönke Jendral |email=uynfo@student.kit.edu |vortragstyp=Bachelorarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2022-08-19 |vo…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Sönke Jendral<br />
|email=uynfo@student.kit.edu<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-08-19<br />
|vortragsmodus=online<br />
|kurzfassung=Advances in computational power have led to increased in interest in machine learning techniques. Sophisticated approaches now solve various prediction problems in the domain of healthcare. Traditionally, machine learning techniques integrate domain knowledge implicitly, by statistically extracting dependencies from their input data. Novel approaches instead integrate domain knowledge from taxonomies as an external component.<br />
<br />
However, these approaches assume the existence of high quality domain knowledge and do not acknowledge issues stemming from low quality domain knowledge. It is thus unclear what low quality domain knowledge in the context of Domain Knowledge Guided<br />
Machine Learning looks like and what its causes are. Further it is not clearly understood what the impact of low quality domain knowledge on the machine learning task is and what steps can be taken to improve the quality in this context.<br />
<br />
In this Thesis we describe low quality domain knowledge and show examples of such knowledge in the context of a sequential prediction task. We further propose methods for identifying low quality domain knowledge in the context of Domain Knowledge Guided Machine Learning and suggest approaches for improving the quality of domain knowledge in this context.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-guided_Load_Disaggregation_in_an_Industrial_Environment&diff=2131Theory-guided Load Disaggregation in an Industrial Environment2022-04-20T00:11:41Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Niels Modry<br />
|email=uulan@student.kit.edu<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-04-22<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=The goal of Load Disaggregation (or Non-intrusive Load Monitoring) is to infer the energy consumption of individual appliances from their aggregated consumption. This facilitates energy savings and efficient energy management, especially in the industrial sector.<br />
<br />
However, previous research showed that Load Disaggregation underperforms in the industrial setting compared to the household setting. Also, the domain knowledge available about industrial processes remains unused.<br />
<br />
The objective of this thesis was to improve load disaggregation algorithms by incorporating domain knowledge in an industrial setting. First, we identified and formalized several domain knowledge types that exist in the industry. Then, we proposed various ways to incorporate them into the Load Disaggregation algorithms, including Theory-Guided Ensembling, Theory-Guided Postprocessing, and Theory-Guided Architecture. Finally, we implemented and evaluated the proposed methods.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-guided_Load_Disaggregation_in_an_Industrial_Environment&diff=2130Theory-guided Load Disaggregation in an Industrial Environment2022-04-20T00:03:54Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Niels Modry |email=uulan@student.kit.edu |vortragstyp=Bachelorarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2022-04-22 |vortr…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Niels Modry<br />
|email=uulan@student.kit.edu<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2022-04-22<br />
|vortragsmodus=in Präsenz<br />
|kurzfassung=The goal of Load Disaggregation (or Non-intrusive Load Monitoring) is to infer the energy consumption of individual appliances from their aggregated consumption. This facilitates energy savings and efficient energy management, especially in the industrial sector.<br />
<br />
However, previous research showed that Load Disaggregation underperforms in the industrial setting, despite all the available domain knowledge about the industrial processes.<br />
<br />
The objective of this thesis was to improve load disaggregation algorithms by incorporating domain knowledge in an industrial setting. First, we identified and formalized several domain knowledge types that exist in the industry. Then, we proposed various ways to incorporate them into the Load Disaggregation algorithms, including Theory-Guided Ensembling, Theory-Guided Postprocessing, and Theory-Guided Architecture. Finally, we implemented and evaluated the proposed methods.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=DomainML:_A_modular_framework_for_domain_knowledge-guided_machine_learning&diff=1806DomainML: A modular framework for domain knowledge-guided machine learning2021-10-05T12:01:21Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Lena Witterauf<br />
|email=lena.emma77@gmail.com<br />
|vortragstyp=Masterarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-10-11 Zusatztermin<br />
|kurzfassung=Standard, data-driven machine learning approaches learn relevant patterns solely from data. In some fields however, learning only from data is not sufficient. A prominent example for this is healthcare, where the problem of data insufficiency for rare diseases is tackled by integrating high-quality domain knowledge into the machine learning process.<br />
<br />
Despite the existing work in the healthcare context, making general observations about the impact of domain knowledge is difficult, as different publications use different knowledge types, prediction tasks and model architectures. It further remains unclear if the findings in healthcare are transferable to other use-cases, as well as how much intellectual effort this requires.<br />
<br />
With this Thesis we introduce DomainML, a modular framework to evaluate the impact of domain knowledge on different data science tasks. We demonstrate the transferability and flexibility of DomainML by applying the concepts from healthcare to a cloud system monitoring. We then observe how domain knowledge impacts the model’s prediction performance across both domains, and suggest how DomainML could further be used to refine both the given domain knowledge as well as the quality of the underlying dataset.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-10-11_Zusatztermin&diff=1805Institutsseminar/2021-10-11 Zusatztermin2021-10-05T11:58:19Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021-10-11T14:00:00.000Z |raum=https://conf.dfn.de/webapp/conference/979148706 }} Zusatztermin am Montag um 14 Uhr.“</p>
<hr />
<div>{{Termin<br />
|datum=2021-10-11T14:00:00.000Z<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}<br />
Zusatztermin am Montag um 14 Uhr.</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=DomainML:_A_modular_framework_for_domain_knowledge-guided_machine_learning&diff=1744DomainML: A modular framework for domain knowledge-guided machine learning2021-07-27T15:27:45Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Lena Witterauf |email=lena.emma77@gmail.com |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2021-07-30 |kurzfass…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Lena Witterauf<br />
|email=lena.emma77@gmail.com<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-07-30<br />
|kurzfassung=Standard, data-driven machine learning approaches learn relevant patterns solely from data. In some fields however, learning only from data is not sufficient. A prominent example for this is healthcare, where the problem of data insufficiency for rare diseases is tackled by integrating high-quality domain knowledge into the machine learning process.<br />
<br />
Despite the existing work in the healthcare context, making general observations about the impact of domain knowledge is difficult, as different publications use different knowledge types, prediction tasks and model architectures. It further remains unclear if the findings in healthcare are transferable to other use-cases, as well as how much intellectual effort this requires.<br />
<br />
With this Thesis we introduce DomainML, a modular framework to evaluate the impact of domain knowledge on different data science tasks. We demonstrate the transferability and flexibility of DomainML by applying the concepts from healthcare to a cloud system monitoring. We then observe how domain knowledge impacts the model’s prediction performance across both domains, and suggest how DomainML could further be used to refine both the given domain knowledge as well as the quality of the underlying dataset.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-07-30&diff=1743Institutsseminar/2021-07-302021-07-27T15:23:56Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/07/30 11:30 |raum=https://conf.dfn.de/webapp/conference/979148706 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/07/30 11:30<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Augmenting_Bandit_Algorithms_with_Domain_Knowledge&diff=1742Augmenting Bandit Algorithms with Domain Knowledge2021-07-21T00:03:10Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Tom George |email=utduu@student.kit.edu |vortragstyp=Masterarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2021-07-23 Zusatzter…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Tom George<br />
|email=utduu@student.kit.edu<br />
|vortragstyp=Masterarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-07-23 Zusatztermin<br />
|kurzfassung=Bandit algorithms are a family of algorithms that efficiently solve sequential decision problems, like monitoring in a cloud computing system, news recommendations or clinical trials. In such problems there is a trade of between exploring new options and exploiting presumably good ones and bandit algorithms provide theoretical guarantees while being practical.<br />
<br />
While some approaches use additional information about the current state of the environment, bandit algorithms tend to ignore domain knowledge that can’t be extracted from data. It is not clear how to incorporate domain knowledge into bandit algorithms and how much improvement this yields. <br />
<br />
In this masters thesis we propose two ways to augment bandit algorithms with domain knowledge: a push approach, which influences the distribution of arms to deal with non-stationarity as well as a group approach, which propagates feedback between similar arms. We conduct synthetic and real world experiments to examine the usefulness of our approaches. Additionally we evaluate the effect of incomplete and incorrect domain knowledge. We show that the group approach helps to reduce exploration time, especially for small number of iterations and plays, and that the push approach outperforms contextual and non-contextual baselines for large context spaces.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Modelling_Dynamical_Systems_using_Transition_Constraints&diff=1739Modelling Dynamical Systems using Transition Constraints2021-07-14T02:19:59Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Florian Leiser<br />
|email=florianleiser.research@gmail.com<br />
|vortragstyp=Masterarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-07-16<br />
|kurzfassung=Despite promising performance of data science approaches in various applications, in industrial research and development the results can be often unsatisfactory due to the costly experiments that lead to small datasets to work with. Theory-guided Data Science (TGDS) can solve the problem insufficient data by incorporating existing industrial domain knowledge with data science approaches. <br />
<br />
In dynamical systems, like gas turbines, transition phases occur after a change in the input control signal. The domain knowledge about the steepness of these transitions can potentially help with the modeling of such systems using the data science approaches. There already exist TGDS approaches that use the information about the limits of the values. However it is currently not clear how to incorporate the information about the steepness of the transitions with them. <br />
<br />
In this thesis, we develop three different TGDS approaches to include these transition constraints in recurrent neural networks (RNNs) to improve the modeling of input-output behavior of dynamical systems. We evaluate the approaches on synthetic and real time series data by varying data availability and different degrees of steepness. We conclude that the TGDS approaches are especially helpful for flat transitions and provide a guideline on how to use the available transition constraints in real world problems. Finally, we discuss the required degree of domain knowledge and intellectual implementation effort of each approach.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Modelling_Dynamical_Systems_using_Transition_Constraints&diff=1738Modelling Dynamical Systems using Transition Constraints2021-07-14T02:18:02Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Florian Leiser |email=florianleiser.research@gmail.com |vortragstyp=Masterarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2021-…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Florian Leiser<br />
|email=florianleiser.research@gmail.com<br />
|vortragstyp=Masterarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-07-16<br />
|kurzfassung=Despite promising performance of data science approaches in various applications, in industrial research and development the results are especially unsatisfactory due to the costly experiments that lead to small datasets to work with. Theory-guided Data Science (TGDS) can solve the problem insufficient data by incorporating existing industrial domain knowledge with data science approaches. <br />
<br />
In dynamical systems, like gas turbines, transition phases occur after a change in the input control signal. The domain knowledge about the steepness of these transitions can potentially help with the modeling of such systems using the data science approaches. There already exist TGDS approaches that use the information about the limits of the values. However it is currently not clear how to incorporate the information about the steepness of the transitions with them. <br />
<br />
In this thesis, we develop three different TGDS approaches to include these transition constraints in recurrent neural networks (RNNs) to improve the modeling of input-output behavior of dynamical systems. We evaluate the approaches on synthetic and real time series data by varying data availability and different degrees of steepness. We conclude that the TGDS approaches are especially helpful for flat transitions and provide a guideline on how to use the available transition constraints in real world problems. Finally, we discuss the required degree of domain knowledge and intellectual implementation effort of each approach.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-07-16&diff=1737Institutsseminar/2021-07-162021-07-14T01:28:08Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/07/16 11:30 |raum=https://conf.dfn.de/webapp/conference/979148706 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/07/16 11:30<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Integrating_Structured_Background_Information_into_Time-Series_Data_Monitoring_of_Complex_Systems&diff=1695Integrating Structured Background Information into Time-Series Data Monitoring of Complex Systems2021-06-11T12:56:40Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Aleksandr Eismont |email=ukqln@student.kit.edu |vortragstyp=Bachelorarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2021-06-18…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Aleksandr Eismont<br />
|email=ukqln@student.kit.edu<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-06-18<br />
|kurzfassung=Monitoring of time series data is increasingly important due to massive data generated by complex systems, such as industrial production lines, meteorological sensor networks, or cloud computing centers. Typical time series monitoring tasks include: future value forecasting, detecting of outliers or computing the dependencies.<br />
<br />
However, the already existing methods for time series monitoring tend to ignore the background information such as relationships between components or process structure that is available for almost any complex system. Such background information gives a context to the time series data, and can potentially improve the performance of time series monitoring tasks.<br />
<br />
In this bachelor thesis, we show how to incorporate structured background information to improve three different time series monitoring tasks. We perform the experiments on the data from the cloud computing center, where we extract background information from system traces. Additionally, we investigate different representations and quality of background information and conclude that its usefulness is independent from a concrete time series monitoring task.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-06-18&diff=1694Institutsseminar/2021-06-182021-06-11T12:54:38Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/06/18 11:30:00 |raum=conf.dfn.de/webapp/conference/979148706 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/06/18 11:30:00<br />
|raum=conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-Guided_Data_Science_for_Battery_Voltage_Prediction:_A_Systematic_Guideline&diff=1614Theory-Guided Data Science for Battery Voltage Prediction: A Systematic Guideline2021-04-06T19:02:15Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Nico Denner<br />
|email=nico.denner@gmx.de<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-04-09 Zusatztermin<br />
|kurzfassung=Purely data-driven Data Science approaches tend to underperform when applied to scientific problems, especially when there is little data available. Theory-guided Data Science (TGDS) incorporates existing problem specific domain knowledge in order to increase the performance of Data Science models. It has already proved to be successful in scientific disciplines like climate science or material research.<br />
<br />
Although there exist many TGDS methods, they are often not comparable with each other, because they were originally applied to different types of problems. Also, it is not clear how much domain knowledge they require. There currently exist no clear guidelines on how to choose the most suitable TGDS method when confronted with a concrete problem.<br />
<br />
Our work is the first one to compare multiple TGDS methods on a time series prediction task. We establish a clear guideline by evaluating the performance and required domain knowledge of each method in the context of lithium-ion battery voltage prediction. As a result, our work could serve as a starting point on how to select the right TGDS method when confronted with a concrete problem.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-Guided_Data_Science_for_Battery_Voltage_Prediction:_A_Systematic_Guideline&diff=1613Theory-Guided Data Science for Battery Voltage Prediction: A Systematic Guideline2021-04-06T19:01:48Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Nico Denner<br />
|email=nico.denner@gmx.de<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-04-09 Zusatztermin<br />
|kurzfassung=Purely data-driven Data Science approaches tend to underperform when applied to scientific problems, especially when there is little data available. Theory-guided Data Science (TGDS) incorporates existing problem specific domain knowledge in order to increase the performance of Data Science models. It has already proved to be successful in scientific disciplines like climate science or material research.<br />
<br />
Although there exist many TGDS methods, they are often not comparable with each other, because they were originally applied to different types of problems. Also, it is not clear how much domain knowledge they require. There currently exist no clear guidelines on how to choose the right TGDS method when confronted with a concrete problem.<br />
<br />
Our work is the first one to compare multiple TGDS methods on a time series prediction task. We establish a clear guideline by evaluating the performance and required domain knowledge of each method in the context of lithium-ion battery voltage prediction. As a result, our work could serve as a starting point on how to select the right TGDS method when confronted with a concrete problem.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-Guided_Data_Science_for_Battery_Voltage_Prediction:_A_Systematic_Guideline&diff=1612Theory-Guided Data Science for Battery Voltage Prediction: A Systematic Guideline2021-04-06T19:00:43Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Nico Denner |email=nico.denner@gmx.de |vortragstyp=Bachelorarbeit |betreuer=Pawel Bielski |termin=Institutsseminar/2021-04-09 Zusatzter…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Nico Denner<br />
|email=nico.denner@gmx.de<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-04-09 Zusatztermin<br />
|kurzfassung=Purely data-driven Data Science approaches tend to underperform when applied to scientific problems, especially when there is little data available. Theory-guided Data Science (TGDS) incorporates existing problem specific domain knowledge in order to increase the performance of Data Science models. It has already proved to be successful in scientific disciplines like climate science or material research.<br />
<br />
Although there exist many TGDS methods, they are often not comparable with each other, because they were originally applied to different types of problems. Also, it is not clear how much domain knowledge they require. There currently exist no clear guidelines on how to choose the right one when confronted with a concrete problem.<br />
<br />
Our work is the first one to compare multiple TGDS methods on a time series prediction task. We establish a clear guideline by evaluating the performance and required domain knowledge of each method in the context of lithium-ion battery voltage prediction. As a result, our work could serve as a starting point on how to select the right TGDS method when confronted with a concrete problem.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-04-09_Zusatztermin&diff=1611Institutsseminar/2021-04-09 Zusatztermin2021-04-06T18:56:15Z<p>If4991: </p>
<hr />
<div>{{Termin<br />
|datum=2021/04/09 11:30<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-04-09_Zusatztermin&diff=1610Institutsseminar/2021-04-09 Zusatztermin2021-04-06T18:54:42Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/04/09 11:30 |raum=Raum 348 (Gebäude 50.34) }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/04/09 11:30<br />
|raum=Raum 348 (Gebäude 50.34)<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Predicting_Dependencies_from_System_Tracing_Data_Instead_of_Computing_Them&diff=1571Predicting Dependencies from System Tracing Data Instead of Computing Them2021-02-23T14:23:47Z<p>If4991: If4991 verschob die Seite Predicting Dependencies from System Tracing Data Instead of Computing Them nach Predicting System Dependencies from Tracing Data Instead of Computing Them</p>
<hr />
<div>#WEITERLEITUNG [[Predicting System Dependencies from Tracing Data Instead of Computing Them]]</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Predicting_System_Dependencies_from_Tracing_Data_Instead_of_Computing_Them&diff=1570Predicting System Dependencies from Tracing Data Instead of Computing Them2021-02-23T14:23:47Z<p>If4991: If4991 verschob die Seite Predicting Dependencies from System Tracing Data Instead of Computing Them nach Predicting System Dependencies from Tracing Data Instead of Computing Them</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Aleksandr Eismont<br />
|email=ukqln@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-02-26 Zusatztermin<br />
|kurzfassung=The concept of Artificial Intelligence for IT Operations combines big data and machine learning methods to replace a broad range of IT operations including availability and performance monitoring of services. In large-scale distributed cloud infrastructures a service is deployed on different separate nodes. As the size of the infrastructure increases in production, the analysis of metrics parameters becomes computationally expensive. We address the problem by proposing a method to predict dependencies between metrics parameters of system components instead of computing them. To predict the dependencies we use time windowing with different aggregation methods and distributed tracing data that contain detailed information for the system execution workflow. In this bachelor thesis, we inspect the different representations of distributed traces from simple counting of events to more complex graph representations. We compare them with each other and evaluate the performance of such methods.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Predicting_System_Dependencies_from_Tracing_Data_Instead_of_Computing_Them&diff=1569Predicting System Dependencies from Tracing Data Instead of Computing Them2021-02-23T14:22:19Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Aleksandr Eismont |email=ukqln@student.kit.edu |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2021-02-26 Zusatz…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Aleksandr Eismont<br />
|email=ukqln@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-02-26 Zusatztermin<br />
|kurzfassung=The concept of Artificial Intelligence for IT Operations combines big data and machine learning methods to replace a broad range of IT operations including availability and performance monitoring of services. In large-scale distributed cloud infrastructures a service is deployed on different separate nodes. As the size of the infrastructure increases in production, the analysis of metrics parameters becomes computationally expensive. We address the problem by proposing a method to predict dependencies between metrics parameters of system components instead of computing them. To predict the dependencies we use time windowing with different aggregation methods and distributed tracing data that contain detailed information for the system execution workflow. In this bachelor thesis, we inspect the different representations of distributed traces from simple counting of events to more complex graph representations. We compare them with each other and evaluate the performance of such methods.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-02-26_Zusatztermin&diff=1568Institutsseminar/2021-02-26 Zusatztermin2021-02-23T14:19:35Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/02/26 11:30:00 |raum=https://conf.dfn.de/webapp/conference/979148706 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/02/26 11:30:00<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Monitoring_Complex_Systems_with_Domain_Knowledge:_Adopting_Contextual_Bandits_to_Tracing_Data&diff=1563Monitoring Complex Systems with Domain Knowledge: Adopting Contextual Bandits to Tracing Data2021-02-09T21:23:42Z<p>If4991: If4991 verschob die Seite Monitoring Complex Systems with Domain Knowledge: Adopting Contextual Bandits to Tracing Data nach Monitoring Complex Systems with Domain Knowledge: Adapting Contextual Bandits to Tracing Data</p>
<hr />
<div>#WEITERLEITUNG [[Monitoring Complex Systems with Domain Knowledge: Adapting Contextual Bandits to Tracing Data]]</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Monitoring_Complex_Systems_with_Domain_Knowledge:_Adapting_Contextual_Bandits_to_Tracing_Data&diff=1562Monitoring Complex Systems with Domain Knowledge: Adapting Contextual Bandits to Tracing Data2021-02-09T21:23:42Z<p>If4991: If4991 verschob die Seite Monitoring Complex Systems with Domain Knowledge: Adopting Contextual Bandits to Tracing Data nach Monitoring Complex Systems with Domain Knowledge: Adapting Contextual Bandits to Tracing Data</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Tom George<br />
|email=utduu@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-02-12 Zusatztermin<br />
|kurzfassung=Monitoring in complex computing systems is crucial to detect malicious states or errors in program execution. Due to the computational complexity, it is not feasible to monitor all data streams in practice. We are interested in monitoring pairs of highly correlated data streams. However we can not compute the measure of correlation for every pair of data streams at each timestep.<br />
<br />
Picking highly correlated pairs, while exploring potentially higher correlated ones is an instance of the exploration / exploitation problem. Bandit algorithms are a family of online learning algorithms that aim to optimize sequential decision making and balance exploration and exploitation. A contextual bandit additional uses contextual information to decide better.<br />
<br />
In our work we want to use a contextual bandit algorithm to keep an overview over highly correlated pairs of data streams. The context in our work contains information about the state of the system, given as execution traces.<br />
A key part of our work is to explore and evaluate different representations of the knowledge encapsulated in traces.<br />
Also we adapt state-of-the-art contextual bandit algorithms to the use case of correlation monitoring.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Monitoring_Complex_Systems_with_Domain_Knowledge:_Adapting_Contextual_Bandits_to_Tracing_Data&diff=1561Monitoring Complex Systems with Domain Knowledge: Adapting Contextual Bandits to Tracing Data2021-02-09T19:38:30Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Tom George |email=utduu@student.kit.edu |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2021-02-12 Zusatztermin…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Tom George<br />
|email=utduu@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-02-12 Zusatztermin<br />
|kurzfassung=Monitoring in complex computing systems is crucial to detect malicious states or errors in program execution. Due to the computational complexity, it is not feasible to monitor all data streams in practice. We are interested in monitoring pairs of highly correlated data streams. However we can not compute the measure of correlation for every pair of data streams at each timestep.<br />
<br />
Picking highly correlated pairs, while exploring potentially higher correlated ones is an instance of the exploration / exploitation problem. Bandit algorithms are a family of online learning algorithms that aim to optimize sequential decision making and balance exploration and exploitation. A contextual bandit additional uses contextual information to decide better.<br />
<br />
In our work we want to use a contextual bandit algorithm to keep an overview over highly correlated pairs of data streams. The context in our work contains information about the state of the system, given as execution traces.<br />
A key part of our work is to explore and evaluate different representations of the knowledge encapsulated in traces.<br />
Also we adapt state-of-the-art contextual bandit algorithms to the use case of correlation monitoring.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-02-12_Zusatztermin&diff=1560Institutsseminar/2021-02-12 Zusatztermin2021-02-09T19:30:20Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/02/12 11:30:00 |raum=https://conf.dfn.de/webapp/conference/979148706 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/02/12 11:30:00<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-02-05_Zusatztermin&diff=1553Institutsseminar/2021-02-05 Zusatztermin2021-02-03T01:46:20Z<p>If4991: </p>
<hr />
<div>{{Termin<br />
|datum=2021/02/05 11:30:00<br />
|raum=https://conf.dfn.de/webapp/conference/979148706<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Modeling_Dynamic_Systems_using_Slope_Constraints:_An_Application_Analysis_of_Gas_Turbines&diff=1552Modeling Dynamic Systems using Slope Constraints: An Application Analysis of Gas Turbines2021-02-02T17:16:18Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Florian Leiser |email=florianleiser.research@gmail.com |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2021-02-0…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Florian Leiser<br />
|email=florianleiser.research@gmail.com<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-02-05 Zusatztermin<br />
|kurzfassung=In energy studies, researchers build models for dynamic systems to predict the produced electrical output precisely. Since experiments are expensive, the researchers rely on simulations of surrogate models. These models use differential equations that can provide decent results but are computationally expensive. Further, transition phases, which occur when an input change results in a delayed change in output, are modeled individually and therefore lacking generalizability.<br />
<br />
Current research includes Data Science approaches that need large amounts of data, which are costly when performing scientific experiments. Theory-Guided Data Science aims to combine Data Science approaches with domain knowledge to reduce the amount of data needed while predicting the output precisely.<br />
<br />
However, even state-of-the-art Theory-Guided Data Science approaches lack the possibility to model the slopes occuring in the transition phases. In this thesis we aim to close this gap by proposing a new loss constraint that represents both transition and stationary phases. Our method is compared with theoretical and Data Science approaches on synthetic and real world data.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Institutsseminar/2021-02-05_Zusatztermin&diff=1551Institutsseminar/2021-02-05 Zusatztermin2021-02-02T17:11:38Z<p>If4991: Die Seite wurde neu angelegt: „{{Termin |datum=2021/02/05 11:30:00 |raum=https://conf.dfn.de/webapp/conference/979160755 }}“</p>
<hr />
<div>{{Termin<br />
|datum=2021/02/05 11:30:00<br />
|raum=https://conf.dfn.de/webapp/conference/979160755<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Theory-Guided_Data_Science_for_Lithium-Ion_Battery_Modeling&diff=1543Theory-Guided Data Science for Lithium-Ion Battery Modeling2021-01-26T10:19:53Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Nico Denner |email=nico.denner@gmx.de |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2021-01-29 |kurzfassung=Li…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Nico Denner<br />
|email=nico.denner@gmx.de<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2021-01-29<br />
|kurzfassung=Lithium-ion batteries are driving innovation in the evolution of electromobility and renewable energy. These complex, dynamic systems require reliable and accurate monitoring through Battery Management Systems to ensure the safety and longevity of battery cells. Therefore an accurate prediction of the battery voltage is essential which is currently realized by so-called Equivalent Circuit (EC) Models. <br />
<br />
Although state-of-the-art approaches deliver good results, they are hard to train due to the high number of variables, lacking the ability to generalize, and need to make many simplifying assumptions. In contrast to theory-based models, purely data-driven approaches require large datasets and are often unable to produce physically consistent results. Theory-Guided Data Science (TGDS) aims at using scientific knowledge to improve the effectiveness of Data Science models in scientific discovery. This concept has been very successful in several domains including climate science and material research. <br />
<br />
Our work is the first one to apply TGDS to battery systems by working together closely with domain experts. We compare the performance of different TGDS approaches against each other as well as against the two baselines using only theory-based EC-Models and black-box Machine Learning models.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Towards_More_Effective_Climate_Similarity_Measures&diff=1434Towards More Effective Climate Similarity Measures2020-09-08T08:52:02Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Pierre Toussing<br />
|email=ulokb@student.kit.edu<br />
|vortragstyp=Bachelorarbeit<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2020-09-11<br />
|kurzfassung=Finding dependencies over large distances — known as teleconnections — is an important task in climate science. To find such teleconnections climate scientists usually use Pearson’s Correlation, but often ignore other available similarity measures, mostly because they are not easily comparable: their values usually have different, sometimes even inverted, ranges and distributions. This makes it difficult to interpret their results. We hypothesize that providing the climate scientists with comparable similarity measures would help them find yet uncaptured dependencies in climate. To achieve this we propose a modular framework to present, compare and combine different similarity measures for time series in the climate-related context. We test our framework on a dataset containing the horizontal component of the wind in order to find dependencies to the region around the equator and validate the results qualitatively with climate scientists.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Towards_More_Effective_Climate_Similarity_Measures&diff=1395Towards More Effective Climate Similarity Measures2020-06-23T11:45:21Z<p>If4991: </p>
<hr />
<div>{{Vortrag<br />
|vortragender=Pierre Toussing<br />
|email=ulokb@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2020-06-26<br />
|kurzfassung=Finding dependencies over large distances — known as teleconnections — is an important task in climate science. To find such teleconnections climate scientists usually use Pearson’s Correlation, but often ignore other available similarity measures, mostly because they are not easily comparable: their values usually have different, sometimes even inverted, ranges and distributions. This makes it difficult to interpret their results. We hypothesize that providing the climate scientists with comparable similarity measures would help them find yet uncaptured dependencies in climate. To achieve this we propose a modular framework to present, compare and combine different similarity measures for time series in the climate-related context. We test our framework on a dataset containing the horizontal component of the wind in order to find dependencies to the region around the equator and validate the results qualitatively with climate scientists.<br />
}}</div>If4991https://sdq.kastel.kit.edu/mediawiki-institutsseminar/index.php?title=Towards_More_Effective_Climate_Similarity_Measures&diff=1394Towards More Effective Climate Similarity Measures2020-06-23T11:42:35Z<p>If4991: Die Seite wurde neu angelegt: „{{Vortrag |vortragender=Pierre Toussing |email=ulokb@student.kit.edu |vortragstyp=Proposal |betreuer=Pawel Bielski |termin=Institutsseminar/2020-06-26 |kurzfas…“</p>
<hr />
<div>{{Vortrag<br />
|vortragender=Pierre Toussing<br />
|email=ulokb@student.kit.edu<br />
|vortragstyp=Proposal<br />
|betreuer=Pawel Bielski<br />
|termin=Institutsseminar/2020-06-26<br />
|kurzfassung=Finding dependencies over large distances — known as teleconnections — is an important task in climate science. To find such teleconnections climate scientists usually use Pearson’s Correlation as a similarity measure, but often ignore other available ones, mostly because they are not easily comparable: their values usually have different, sometimes even inverted, ranges and distributions. This makes it difficult to interpret their results. We hypothesize that providing the climate scientists with comparable similarity measures would help them find yet uncaptured dependencies in climate. To achieve this we propose a modular framework to present, compare and combine different similarity measures for time series in the climate-related context. We test our framework on a dataset containing the horizontal component of the wind in order to find dependencies to the region around the equator and validate the results qualitatively with climate scientists.<br />
}}</div>If4991