Review of dependency estimation with focus on data efficiency: Unterschied zwischen den Versionen

Aus SDQ-Institutsseminar
Keine Bearbeitungszusammenfassung
Keine Bearbeitungszusammenfassung
Zeile 5: Zeile 5:
|betreuer=Bela Böhnke
|betreuer=Bela Böhnke
|termin=Institutsseminar/2022-01-14 Zusatztermin
|termin=Institutsseminar/2022-01-14 Zusatztermin
|kurzfassung=In our data-driven world, where tons of data are collected, dependency estimation is important to get more insights into our data. Many dependency estimation algorithms are hard to use in a real-world setting. In this study, I will do a comparison of different state-of-the-art dependency estimation algorithms. For comparison, a list of different criteria is used and the focus of this study is on data efficiency and uncertainty of the dependency estimation algorithms. The comparison includes a theoretical analysis and a variety of different experiments with an implementation of the dependency estimation algorithm that performed well in the theoretical analysis.
|kurzfassung=In our data driven world, where large amounts of data are collected in all kind of environments, data analysis rises in importance. How different variables influence each other is an major part of knowledge discovery and allow strategic decisions based on this knowledge. Therefor high quality dependency estimation should be accessible  to a variety of people. Many dependency estimation algorithms are difficult to use in a real-world setting. Many dependency estimation algorithms need very large data sets to return a good estimation. In practice gathering this amount of data may be costly. Especially when the data is collected in experiments with high cost for materials or infrastructure needed for the experiment. In this study different state-of-the-art dependency estimation algorithms are compared by a list of different criteria. The focus of this study is on data efficiency and uncertainty of the dependency estimation algorithms. The comparison includes a theoretical analysis and different experiments with dependency estimation algorithms that performed well in the theoretical analysis.
}}
}}

Version vom 5. Januar 2022, 15:11 Uhr

Vortragende(r) Maximilian Georg
Vortragstyp Proposal
Betreuer(in) Bela Böhnke
Termin [[Institutsseminar/2022-01-14 Zusatztermin|]]
Vortragsmodus
Kurzfassung In our data driven world, where large amounts of data are collected in all kind of environments, data analysis rises in importance. How different variables influence each other is an major part of knowledge discovery and allow strategic decisions based on this knowledge. Therefor high quality dependency estimation should be accessible to a variety of people. Many dependency estimation algorithms are difficult to use in a real-world setting. Many dependency estimation algorithms need very large data sets to return a good estimation. In practice gathering this amount of data may be costly. Especially when the data is collected in experiments with high cost for materials or infrastructure needed for the experiment. In this study different state-of-the-art dependency estimation algorithms are compared by a list of different criteria. The focus of this study is on data efficiency and uncertainty of the dependency estimation algorithms. The comparison includes a theoretical analysis and different experiments with dependency estimation algorithms that performed well in the theoretical analysis.