This commit is contained in:
Annkathrin 2021-08-19 14:20:17 +02:00
parent 70b6fcb8d9
commit 81a06c74a5
3 changed files with 16 additions and 11 deletions

17
IBC.log
View file

@ -1,4 +1,4 @@
This is LuaHBTeX, Version 1.13.0 (TeX Live 2021) (format=lualatex 2021.8.7) 12 AUG 2021 11:58
This is LuaHBTeX, Version 1.13.0 (TeX Live 2021) (format=lualatex 2021.8.7) 19 AUG 2021 00:49
restricted system commands enabled.
**IBC.tex
(./IBC.tex
@ -1040,20 +1040,23 @@ LaTeX Font Info: Font shape `TU/lmtt/bx/n' in size <10> not available
] [16
] [17
]
Overfull \vbox (30.55852pt too high) detected at line 182
[]
[17
]
<Ergebnis_ibc.png, id=216, 547.04375pt x 492.84125pt>
File: Ergebnis_ibc.png Graphic file (type png)
<use Ergebnis_ibc.png>
Package luatex.def Info: Ergebnis_ibc.png used on input line 190.
Package luatex.def Info: Ergebnis_ibc.png used on input line 192.
(luatex.def) Requested size: 196.9356pt x 177.4227pt.
[18
<./Ergebnis_ibc.png>] [19
]
[20
] [20
]
\tf@nav=\write6
@ -1077,7 +1080,7 @@ Here is how much of LuaTeX's memory you used:
38 hlist, 7 vlist, 2 rule, 6 local_par, 111 glue, 10 kern, 27 penalty, 302 gl
yph, 79 attribute, 61 glue_spec, 79 attribute_list, 4 write, 32 pdf_colorstack n
odes
avail lists: 2:284,3:147,4:250,5:64,6:61,7:2144,8:50,9:243,10:12,11:6
avail lists: 2:284,3:148,4:250,5:64,6:61,7:2642,8:50,9:246,10:12,11:6
41767 multiletter control sequences out of 65536+600000
64 fonts using 5902119 bytes
131i,11n,125p,406b,435s stack positions out of 5000i,500n,10000p,200000b,80000s
@ -1086,7 +1089,7 @@ odes
f></usr/local/texlive/2021/texmf-dist/fonts/opentype/public/lm/lmmonolt10-bold.o
tf></usr/local/texlive/2021/texmf-dist/fonts/opentype/public/lm/lmmono10-regular
.otf>
Output written on IBC.pdf (20 pages, 124882 bytes).
Output written on IBC.pdf (20 pages, 125391 bytes).
PDF statistics: 270 PDF objects out of 1000 (max. 8388607)
222 compressed objects within 3 object streams

BIN
IBC.pdf

Binary file not shown.

10
IBC.tex
View file

@ -164,18 +164,20 @@ Beispiel (Profiling?)
\end{itemize}
\end{frame}
\section{Bias Bingo mit / Bias in iBorderCtrl (20min)(interaktiv?)}
\begin{frame}{Bias in iBorderCtrl}
\begin{itemize}
\item Measurement Bias: Wie wird entschieden ob jemand gelogen hat o.Ä.?
\item Measurement Bias: Schauspieler*innen statt echte Situationen
\item Omitted Variable Bias: Nervosität durch Stress beim Grenzübergang
\item Representation Bias: fehlende Diversität race, gender, neurdiversity, disability, health, scars
\item Aggregation Bias: s. representation bias, teilweiser Versuch der Gegensteuerung
\item Sampling Bias: ungleiche Datenmengen
\item Overfitting
\item ...
\item Overfitting: 73 Prozent in Testdaten vs 93 Prozent in Trainingsdaten, das impliziert Overfitting ist wahrscheinlich
\item Historical Bias: höhere Wahrscheinlichkeit zB des Drogenschmuggels bestimmter Gruppen (trifft erst bei Benutzung zu, weil Testdaten zu klein)
\item Algorythmischer Bias: zu wenig Einblick
\item Evaluations Bias: Testbedingungen entsprechen nicht den Einsatzbedingungen, zB Licht, Diversität
\item Bias durch Rückkopplung:
\end{itemize}
\end{frame}