TY - GEN
T1 - Evaluating Deep Learning Biases Based on Grey-Box Testing Results
AU - Jenny Li, J.
AU - Silva, Thayssa
AU - Franke, Mira
AU - Hai, Moushume
AU - Morreale, Patricia
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - The very exciting and promising approaches of deep learning are immensely successful in processing large real world data sets, such as image recognition, speech recognition, and language translation. However, much research discovered that it has biases that arise in the design, production, deployment, and use of AI/ML technologies. In this paper, we first explain mathematically the causes of biases and then propose a way to evaluate biases based on testing results of neurons and auto-encoders in deep learning. Our interpretation views each neuron or autoencoder as an approximation of similarity measurement, of which grey-box testing results can be used to measure biases and finding ways to reduce them. We argue that monitoring deep learning network structures and parameters is an effective way to catch the sources of biases in deep learning.
AB - The very exciting and promising approaches of deep learning are immensely successful in processing large real world data sets, such as image recognition, speech recognition, and language translation. However, much research discovered that it has biases that arise in the design, production, deployment, and use of AI/ML technologies. In this paper, we first explain mathematically the causes of biases and then propose a way to evaluate biases based on testing results of neurons and auto-encoders in deep learning. Our interpretation views each neuron or autoencoder as an approximation of similarity measurement, of which grey-box testing results can be used to measure biases and finding ways to reduce them. We argue that monitoring deep learning network structures and parameters is an effective way to catch the sources of biases in deep learning.
KW - Bias measurement
KW - Deep learning
KW - Deep learning evaluation
KW - Mathematical interpretation
KW - Neural Networks
UR - http://www.scopus.com/inward/record.url?scp=85090171668&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-55180-3_48
DO - 10.1007/978-3-030-55180-3_48
M3 - Conference contribution
AN - SCOPUS:85090171668
SN - 9783030551797
T3 - Advances in Intelligent Systems and Computing
SP - 641
EP - 651
BT - Intelligent Systems and Applications - Proceedings of the 2020 Intelligent Systems Conference IntelliSys Volume 1
A2 - Arai, Kohei
A2 - Kapoor, Supriya
A2 - Bhatia, Rahul
PB - Springer
T2 - Intelligent Systems Conference, IntelliSys 2020
Y2 - 3 September 2020 through 4 September 2020
ER -