Skip to main content

AI bias

[PLACEHOLDER]

Objectivity can be challenging. Machines are not shielded from this by any means.

As humans we are leveraging technology in a progressive manner, making our lives easier, becoming an extension of ourselves in our day-to-day existence. By trying to simplify things, complexity creeps in (it is ironic, I know!), and as we try to create something for the good, very often it can be used to go against the principle upon why they were created in the first place.

 

Photo by Pavel Danilyuk from Pexel

Therefore, as creators of this technology, usually with the best intentions, it is understandable when model predictions become susceptible to bias.

Those in charge of building the models need to be aware of the common human biases that will find their way into the data used, allowing them to take proactive mitigation steps.

To help remove those biases then we need to have structure. A framework that allows us to build AI systems in an ethical manner that can benefit our communities. AI ethics is that “set of guidelines that advise on the design and outcomes of Artificial Intelligence (AI)”.

AI and bias

Algorithm bias “describes systematic and repeatable errors in a computer system that create unfair outcomes” - Wikipedia.

Putting close attention to the algorithms, and the data, selected for the implementation of the solution “X” allows to identify, prevent and/or mitigate real concerns, such as systematic and unfair discrimination.
AI bias can happen due to cognitive biases, which is an unconscious error in thinking that leads you to misinterpret information from the world around you, affecting the rationality and accuracy of decisions and judgement.

The teams in charge of training the models can be ingesting those biases through the data collected.
Note: The use of incomplete data for data modelling training is another potential way to have AI bias as your outcome.

The use of incomplete data for data modeling training is another potential way to have AI bias as your outcome


AI bias types in ML 

Here are a group of bias types that can be found in data (Go to the Google Developers link in our references section for more details and examples): 

Automation bias Selection bias Group Attribution bias Implicit bias


Reporting bias

When the frequency of events, properties, and/or outcomes captured in a data set does not accurately reflect their real-world frequency.

Automation bias

a tendency to favour results generated by automated systems over those generated by non-automated systems.

Selection bias

Occurs if a data set's examples are chosen in a way that is not reflective of their real-world distribution.

Group Attribution bias

A tendency to generalize what is true of individuals to an entire group to which they belong.

Implicit bias

occurs when assumptions are made based on one's own mental models and personal experiences that do not necessarily apply more generally.


Reference

 

Trending posts

Goal setting frameworks for Product Management - OKR and HOSKR

As a business analyst and product manager we often use various frameworks to synthesize and organize our product ideas and goals. I think of frameworks as tools in our product management tool kit which we use depending on the task at hand.  And speaking of goals, OKR is a very popular framework that I often use to set the goals for the products I am managing. However recently I participated the #ProductCon conference hosted by Product School  and I stumbled upon one of the talks in which Rapha Cohen, the CPO at Google Waze introduced a more effective framework for setting product goals. The framework is called HOSKR.  In this post I'll describe both the OKR and HOSKR frameworks in more details using examples. I hope this will provide you, our readers, more practical insights on how to effectively use these frameworks to set your product goals.  OKR OKR stands for O bjectives and K ey R esults. If you are reading this post then you are on our Beolle blog and I am going to use one o

Big O notation

 Big O notation is a mathematical notation used for describing the limiting behaviour of a function. As a developer (programmer), you could use Big O to understand the performance of an algorithm. By using this concept a developer can evaluate a function’s run-time based on the input(s) passed.  A developer can also use it to compare algorithms' efficiency within the same domain. This comparison can well be for evaluating the performance. The algorithm that yields the lower Big O value is identified as the optimal one. Photo by Anni Roenkae from Pexels What is the "O" in Big O? “The letter O is used because the growth rate of  a function is also referred to as the order of the function” - wikipedia ( https://en.wikipedia.org/wiki/Big_O_notation ). In the list below you can see the various functions: Notation Name O ( 1 ) Constant O ( log

Your platform as a positive force in the world

If you are managing a platform, even if it is small, then you are leveraging one of the elements and popular channels used to build digital communities. These platforms, and channels, can come in the shape of websites, blogs, mobile apps, podcasts, vlogs, social media, or a combination of the ones mentioned. Photo by Tatiana Syrikova from Pexels   There are many reasons for building a community, such as sharing knowledge, promoting businesses, branding, the need for engaging with others, etc. However as Spiderman’s Uncle Ben said: “With great power comes great responsibility.” If you like to build better communities with more collaboration and less silos then how would you do it? How can communities come together to drive a “ positive force ”? The current world is mired in conflicts. How long are we willing to go on in such disarray? All the continents are in some kind of “a demonstration of partisanship, intolerance and selfishness”. There are nations in the Americas, Africa, Middl

Adobe Summit 2022 - Experience Platform

 Our hero image features many air balloons flying, scattered in a blue sky. Each of them has characteristics in their designs that are “part of a whole”, which in this case is an exhibition that is part of a festival. Now, let us leverage this analogy, mapping it to the elements that are relevant for this post’s agenda. Think about the pieces of data scattered through different systems, which combined (behaviour and PII data) becomes part of a whole: an enriched consumer profile that can provide you a “360 view”. This allows corporations to find opportunities, and turn them into value: “enabling ways to engage with that customer(s) to meet a cluster of business goals.” Photo by Lad Fury from Pexels   The objectives for this article are: To provide our takeaways from two (2) Adobe summit 2022 sessions, related to the Adobe Experience Platform and Real-time CDP (Customer Data Platform). Those sessions being S401 (AEP a modern foundation - by Klaasjan Tukler) and  S408 (DMP vs CDP:

This blog uses cookies to improve your browsing experience. Simple analytics might be in place for pageviews purposes. They are harmless and never personally identify you.

Agreed