Scribd
Upload
58 views

Unit1 Intro AI

Unit1 Intro AI

Uploaded by

vibhay vibhay
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
58 views

Unit1 Intro AI

Unit1 Intro AI

Uploaded by

vibhay vibhay
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 40

Neural Network and

Deep Learning

Class X-AI
Applications of Deep learning

• Fraud Detection.
• Autonomous Cars.
• Virtual Assistants.
• Facial Recognition Systems.
Neural networks
• Neural networks are a set of
algorithms, modelled loosely after the
human brain, that are designed to
recognize patterns.
• Neural networks help us cluster and
classify.
Clustering

• Clustering or grouping is the detection


of similarities.
• Deep learning does not require labels
to detect similarities.
• Learning without labels is called
unsupervised learning.
Activity 4:Can you create a
decision tree for someone
trying to enter a room!
Machine learning Algorithms
• Algorithms are a set of standard
instructions for a computer device on
how to handle data.
• In machine learning data itself creates
model.
• with more and more data the model
become better and accurate.
Decision Tree
• A decision tree is drawn upside down
with its root at the top.
Dimensionality Reduction
• Dimensionality: the number of inputs
present in a given data set.
• Technique used to remove non-useful
or redundant data.
Instance based Algorithm
•Instance based algorithm or Lazy –learners use training data to
compare with the new data set or test data set.
•Categorize the new data points based on the similarities of the
trained data .
Regularization to avoid over
fitting
• Overfitting is a problem in machine learning with
limited data set.
• regularization is a technique to modify model to
avoid the problem of Overfitting.
Rule based Machine learning
• Relational rules to describe data.
• with the set of input , output is
described as a relation with the input
as y=f(x).
• Training a machine learning algorithm
include 3 steps:
– Representation
– Evaluation
– Optimization
• Training a machine learning algorithm
include 3 steps:
– Representation : interprets data
– Evaluation : Identify which model produce
accurate prediction.
– Optimization : understanding data is
critical to your success. If you create a
model based on faulty data your
predictions will be inaccurate
Examples of Structured Data
 Sensor data
Server and log data
Financial Data
Weather data
Website/E-commerce data
Examples of unstructured
Data
• Undocumented data
• Social media
• Mobile data
• Surveillance data
Domains of AI
• Three domains
– Data science
– Computer vision
– Natural language processing
• Data collection is the process of
identifying various source of
data( Structured and unstructured) and
collect them and prepare to label it.
Monica Rogati’s data science hierarchy of needs
Big data and AI
• AI depends on Big data structure for
success.
• Big Data is a collection of data that
is huge in volume, yet growing
exponentially with time.
Augmented Intelligence
• Combine power of human intuition with machine
intelligence.
• AI system need to learn from data as well as from
human, in order to be able to fulfil its function.
Computer Vision
• CV is a subset of AI that lets machine
see and extract meaning from pixel in
an image.
• In biological term, CV aims to mirror
how human vision works and interpret
things we see.
How Computer Vision
works?
Difference between CV and
image processing
• CV: Interpreting the content of the
image.
• Image processing: the process of
creating a new image from an already
existing image, either simplifying or
enhancing the content
• Google glass:
https://www.youtube.com/watch?v=4EvN
xWhskf8

• Magic of Holo Lens Mixed Reality


• https://www.youtube.com/watch?v=auJ
JrHgG9Mc
Natural Language Processing
(NLP)
• Algorithm that understand and interpret
speech and text which is natural language
data.
• NLP is the heart of modern AI input/output
system.
• The ability to understand NL include two
capabilities:
– Comprehend unstructured data.
– Mimic human conversation
• The goal of NLU system( Natural
Language Understanding) is to figure
out the meaning of language inputs.
(words,Sentences,Stories etc).
• System use the combination of three
Kind of Information
– Pragmatics
– Semantics
– Syntax
Ethics in AI

• https://www.youtube.com/watch?v=vg
UWKXVvO9Q

• Ethics are defined as a set of moral


principals , guiding actions of an
individual or a group, helping to
determine what is good or right.
Two challenges
• How does the team of developer
determine what is good or right
outcome and for whom?
• AI is autonomous, self-learning and
self –improving technology. This
means it does most of its decision-
making itself based on its own analysis
of data
Summary
•
–AI Ethics
Basics of Ethics:
– Bias,Prejudice and Fairness
– Who is accountable
– Transparency, interpretability and Explainability
• Action of AI:
– Safety
– Human-AI interaction
– Cyber security and wrong intentional use
– Data privacy and control
• AI impact:
– Job losses and unemployment
– Civil Rights-Robot Rights
– Human-Human interaction changes
– Economy
Data Privacy

• The world of Artificial Intelligence


revolves around Data.
• Where do we collect data from?
• Why do we need to collect data?
• Are we okay with sharing our data with
the external world?
• Why do these apps collect data?
• We need to understand that the data which is
collected by various applications is ethical as the
Smartphone users agree to it (by clicking on allow
when it asks for permission and by agreeing to all
the terms and conditions).

• But at the same time if one does not want to share


his/her data with anyone, he/she can opt for
alternative applications which are of similar usage
and keep your data private.
AI Bias
• Everyone has a bias of their own.
• Biases are not negative all the time.
Sometimes.
• it is required to have a bias to control a
situation and keep things working.
• we cannot expect a machine to have any
biases of its own.
• Any bias can transfer from the developer
to the machine while the algorithm is
being developed.
AI access
• AI is a budding technology.
• Not everyone has the opportunity to access
it.
• AI creates unemployment and ethical
issues.
– AI machines will replace those who work
as labourers.
– People with little or no skill will be left
without job.
– Who keep up their skill will flourish.
Potential Harms from
Automated Decision -Making

You might also like