Explainable: By-Vikram Rajpurohit 23ADR185 Kongu Engineering College 2023-2027

XAI
EXPLAINABLE AI
By-
Vikram Rajpurohit
23ADR185
Kongu Engineering College
2023-2027
Third
AI wave of AI
Symbolic Statistical Explainable
AI AI models for
Statistical AI
Logic rules represent specific domains Systems construct
Knowledge. training on big data. explanatory models.
No learning capability No contextual Systems learn and

and poor handling of capability and minimal reason with new tests
updated data. explainability. and situations.
Factors driving rapid advancement of AI
GPU Data Cloud New

Black Box
AI
Black box AI refers to complex models, whose decision-
making processes are not easily understandable by humans.
The lack of transparency in these systems can raise
concerns about accountability and trust. Black box AI
models arrive at conclusions or decisions without providing
any explanations as to how they were reached.
Confusion with today's AI
Black Box
• Why did you do that?
• Why did you not do that?
• When do you succeed or fail?
• How do I correct an error?
What is
XAI ?Explainable Al or XAI. The goal of XAI is to provide
XAI is a Al system that explains their decision making which is
referred as
verifiable explanations of how machine learning systems makes
decisions and let humans to be in the loop.
There are two ways to provide explainable AI
• Easy-to-Understand Models:
Use AI models that are naturally easy to understand, like
decision trees or models that find similarities.
• Find New Ways for Complex Models:

For more complicated models, like neural networks, come up
with new methods to explain how they make decisions. This helps
make even the complex models clearer to people.
EXPLAINABLE AI – Three Key Pillars
Reasonabl
Reasonable
The ability to understand
eAI the reasoning behind
AI each individual prediction
EXPLAINABLE Traceabl
Traceable
EXPLAINABLE The ability to trace
AI e AI prediction process from
AI
AI logic of math algo to nature
of data.
Understandab
Understandab
le The ability to fully
leAI
AI understand the model
upon which the AI
Why XAI?
1.Interpretability and Transparency:
• XAI provides clear and interpretable explanations for AI decisions,
promoting transparency.
• Black-box models, while powerful, often lack transparency, making it
challenging to interpret their outputs.
2. Trust and Accountability:
• XAI builds trust by enabling users to comprehend and validate the decision
making process.
• Black-box models may erode trust due to their opaque nature, especially in
critical applications.
3. Regulatory Compliance:
• XAI aligns with regulations that mandate transparency and fairness in AI systems.
• Black-box models may face challenges in complying with such regulations,
potentially leading to legal and ethical issues.
Process of XAI
• The significant enabler of explainable Al is Interpretability. It
collaborates between human and artificial intelligence.
• Interpretability is a degree to which a human can understand the cause of

a decision.
• It strengthens trust and transparency, explains decisions,

fulfil regulatory requirements, and improve models.
• The stages of Al explainability is categorized into pre-

modelling, explainable modelling and post-modelling. They focus
on explainability at the dataset stage and during model development.
Applications of XAI
1.Healthcare: XAI provides transparent explanations for medical
diagnoses and treatment plans, ensuring trust and understanding in
healthcare decisions.
2.Finance:XAI enhances transparency in credit scoring and fraud
detection, promoting fairness and accountability in financial practices.
3.Education:XAI supports personalized learning and mitigates biases

in educational algorithms, ensuring fair and understandable
recommendations.
4.Research and Development: XAI aids in scientific discovery
and biomedical research by providing interpretable insights into
model predictions.
5.IoT and Smart Systems: XAI explains decisions in IoT devices and s
mart systems, enhancing user understanding and trust
in these technologies.
Conclusio
n guide that helps us understand how smart computer systems
• In summary, Explainable Artificial Intelligence (XAI) is like a
make decisions.
• It's important because it makes things like medical diagnoses,
financial decisions, and self-driving cars more transparent and easier
to trust.
• Even though there are some challenges, researchers are working to

make these systems both powerful and easy for us to understand.
• XAI is making sure that as technology gets smarter, it also

stays fair, accountable, and something we can rely on.
T HA NK
YO U

Explainable: By-Vikram Rajpurohit 23ADR185 Kongu Engineering College 2023-2027

Uploaded by

Copyright:

Available Formats

Explainable: By-Vikram Rajpurohit 23ADR185 Kongu Engineering College 2023-2027

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Explainable: By-Vikram Rajpurohit 23ADR185 Kongu Engineering College 2023-2027

Uploaded by

Copyright:

Available Formats

XAI

No learning capability No contextual Systems learn and

Factors driving rapid advancement of AI

GPU Data Cloud New

• Find New Ways for Complex Models:

• Interpretability is a degree to which a human can understand the cause of

• It strengthens trust and transparency, explains decisions,

• The stages of Al explainability is categorized into pre-

3.Education:XAI supports personalized learning and mitigates biases

• Even though there are some challenges, researchers are working to

• XAI is making sure that as technology gets smarter, it also

You might also like