MindMap Gallery Ethical reflections on algorithmic discrimination
Research on machine learning fairness, algorithmic discrimination caused by pre-existing biases in data, and algorithmic discrimination caused by sampling bias and different weights of data in algorithmic decision-making.
Edited at 2023-08-01 16:39:01Avatar 3 centers on the Sully family, showcasing the internal rift caused by the sacrifice of their eldest son, and their alliance with other tribes on Pandora against the external conflict of the Ashbringers, who adhere to the philosophy of fire and are allied with humans. It explores the grand themes of family, faith, and survival.
This article discusses the Easter eggs and homages in Zootopia 2 that you may have discovered. The main content includes: character and archetype Easter eggs, cinematic universe crossover Easter eggs, animal ecology and behavior references, symbol and metaphor Easter eggs, social satire and brand allusions, and emotional storylines and sequel foreshadowing.
[Zootopia Character Relationship Chart] The idealistic rabbit police officer Judy and the cynical fox conman Nick form a charmingly contrasting duo, rising from street hustlers to become Zootopia police officers!
Avatar 3 centers on the Sully family, showcasing the internal rift caused by the sacrifice of their eldest son, and their alliance with other tribes on Pandora against the external conflict of the Ashbringers, who adhere to the philosophy of fire and are allied with humans. It explores the grand themes of family, faith, and survival.
This article discusses the Easter eggs and homages in Zootopia 2 that you may have discovered. The main content includes: character and archetype Easter eggs, cinematic universe crossover Easter eggs, animal ecology and behavior references, symbol and metaphor Easter eggs, social satire and brand allusions, and emotional storylines and sequel foreshadowing.
[Zootopia Character Relationship Chart] The idealistic rabbit police officer Judy and the cynical fox conman Nick form a charmingly contrasting duo, rising from street hustlers to become Zootopia police officers!
Ethical reflections on algorithmic discrimination
Summary
Ethical Issues of Legal Discrimination
Algorithmic fairness
Algorithmic identity stigmatization
Algorithmic privacy
reason
Algorithmic discrimination caused by pre-existing biases in data
Using an algorithm itself may be a form of discrimination
Sampling bias and different weights of data in algorithmic decision-making lead to algorithmic discrimination
Avoid algorithmic discrimination
Technical approach
philosophical approach
Approach to the rule of law
Key words
algorithmic discrimination
Algorithmic fairness
Algorithmic identity stigmatization
Algorithmic discrimination and its emergence
Manifestations
explicit discrimination
Direct reliance on factors related to protected groups
Algorithms use characteristics such as race as a factor in decision-making
Features
The subjective discriminatory intention (i.e. the intention of the will) is more obvious
implicit discrimination
Bias embedded in software instructions
Biased data sets and tool usage
Choose with purpose
training data
Label
feature
Features
Discriminatory intentions are hidden deeper
differential impact
Characteristics that only have a differential impact
Algorithms do not use characteristics of protected groups, but decision outcomes are still skewed in favor of minority groups
Features
Neutrality in form and discrimination in practice
category
Does discrimination based on algorithms involve identity rights?
Prejudice against a certain identity
Algorithmic discrimination related to identity rights
religion
nation, race
gender
gender
Algorithmic discrimination without identity rights
Big data familiarity
price discrimination
Why did it come about
Biased training data leads to feedback loops
Using algorithms can itself be a form of discrimination
Sampling bias of the input data and poor weight settings result in
Ethical issues related to algorithmic discrimination
Fairness: individual fairness and group fairness
two questions
Can fairness be quantified and formalized?
Which equity theory is appropriate to choose, if it can be quantified?
For the first question
The answer is yes
From the current research results
Data mining and algorithmic fairness to prevent discrimination become new research paradigms
For the second question
fair quantification form
group fairness
is called
statistical parity
feature
The proportion of people receiving positive or negative classification is the same as the entire population
Aim to treat all groups equally
Requires decision outcomes to be equally proportional
protected group
non-protected group
individual fairness
feature
Treat similar individuals equally
Measure the similarity between individuals on a specific task
Prevent individuals from being discriminated against
Belonging to a group
stigmatization of identity
two aspects
Algorithmic discrimination concerns what algorithmic identity an individual is assigned to
Individuals suffer a double cumulative disadvantage
Once labeled as being easily discriminated against
Approaches to Solving the Ethical Issues of Algorithmic Discrimination
Technical approach
mainly focus on
Technology research and development by professional technicians
Prevent algorithmic discrimination
Achieve algorithmic fairness
mainly include
"beforehand"
Data preprocessing
algorithm design
"afterwards"
Post-event supervision
post audit
philosophical approach
Argument
some kind of "fairness" theory
rationality
limitation
possibility of realization
Approach to the rule of law
Legally regulate algorithms such as data sources and transparency
Main challenges
How to legally determine the standard of algorithmic discrimination
Conclusion
Research on foreign algorithmic discrimination
Instructive for Chinese scholars
Cognitive Algorithmic Discrimination
Possible solution paths
Research on China is relatively weak
Neglect of algorithmic discrimination and algorithmic fairness at the policy and regulatory levels
At the industry and technical levels, my country’s research on algorithmic fairness needs to be strengthened
Cultivation of ethical awareness of scientific and technological personnel