[go: up one dir, main page]
Scribd
Upload
38 views54 pages

Class1 Code of Conduct Motivation Batch2

The document outlines the syllabus, materials, and evaluation scheme for an artificial intelligence course, including topics like search heuristics, game theory, learning theory, and supervised and unsupervised learning methods. It also provides examples of applications of AI such as games, smart cities, medical robots, and blockchain technology. The course aims to teach students the basics of AI and how to apply AI techniques to solve real-world problems.

Uploaded by

Bhuvnesh Sharma
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
38 views54 pages

Class1 Code of Conduct Motivation Batch2

The document outlines the syllabus, materials, and evaluation scheme for an artificial intelligence course, including topics like search heuristics, game theory, learning theory, and supervised and unsupervised learning methods. It also provides examples of applications of AI such as games, smart cities, medical robots, and blockchain technology. The course aims to teach students the basics of AI and how to apply AI techniques to solve real-world problems.

Uploaded by

Bhuvnesh Sharma
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 54

Class1: Artificial and Computational

!1
Intelligence

Yamuna Prasad
CSE, IIT Jammu
Outline
• Credit Structure & Syllabus
• Reference Books and Materials
• Online Programming Editors & Compilers
• Evaluation Scheme: Assignments, quizzes, MidTerm &
EndTerm
• Hardware Requirements
• Teaching Assistants
• Time Table
• Code of Conduct
• Few Applications and Motivation
• Any general query

Note: Most of the images are from


Google Search engines
Credit Structure & Syllabus
• L-T-P (3-0-0) As per IIT Jammu Curriculum
• 3 hours of lecture, 0 hour for tutorial and 0 hour lab per
week
• 2 hours of classes per week and sometimes 1 hour of
Tutorial per week
• Usually 42 hours of lecture in a semester

• Course Codes:
• Theory (lecture): COL001P1M
• Type: Core for MTech for Executives (CEAC)
Course Objective
• Artificial intelligence and basics of learning
methods
• Learning Approaches to Solve Complex Problems
involving Games, Forecasting, Data Enhancement
etc.
• Pre-requisite: Python programming (simple
constructs, syntax, functions, recursions, file
handling and exception handling, basic concepts of
modularity and object- oriented programming &
graphics etc.)
Course Outcome
• A Learn the basics and applications of artificial
intelligence, learning basics, search heuristics and
categorise various problem domains, basic
knowledge representation and reasoning methods.
• Apply AI techniques to real-world problems to
develop intelligent systems
• Select appropriate methods from a range of
techniques
Course Contents: Total 42 hours
• Basics of Artificial Intelligence [5 hours]
• Search Heuristics and Goal Trees [12 hours]
• Game Theory and Logic [08 hours]
• Learning Theory [04 hours]
• Supervised and Unsupervised Methods [08 hours]
• Problem Solving [05 hours]

• Time Table: Saturday (11-1PM)


• Discussion Hours (It can be scheduled in
Tutorial sessions)
Course Materials
1. S. Russel and P. Norvig, “Artificial Intelligence – A Modern
Approach”, SecondEdition, Pearson Education
2. David Poole, Alan Mackworth, Randy Goebel,
”Computational Intelligence : a logical approach”, Oxford
University Press.
3. G. Luger, “Artificial Intelligence: Structures and Strategies
for complex problem solving”, Fourth Edition, Pearson
Education.
4. J. Nilsson, “Artificial Intelligence: A new Synthesis”, Elsevier
Publishers
5. Alpaydin E., Introduction to Machine Learning, MIT Press
(2010).
6. Reference materials:
https://ocw.mit.edu/courses/6-034-artificial-intelligence-
fall-2010/
Course Materials: Programming
1. Notebooks & Environments: Google Colab
(https://colab.research.google.com/), Jupyter,
Anaconda, spider etc. (Python3/Python2)
2. Online study materials (from coursera, geeks,
Edureka, Towardsdatascience, Medium,… etc)
Evaluation Scheme
• Assignments:
• Theory + Lab: 10 Assignments [20 Marks]
• Quizzes:
• Theory + Lab: 4 best out of 5 quizzes [20 Marks]

• Exams:
• Theory: MidTerm [30 Marks]; EndTerm [30
Marks]
Submission of Assignments:
• Name of the Assignment Solution:
• FirstName_EnrollmentID_GroupID_Assignment
Name
• Examples: for assignment1, student name:
Abhinish Kumar, 2021UCH0001, Group 1
• The assignment solution file name will be:
• abhinish_2021uch0001_G1_assignment1
Hardware & Software
• Processor: i3 onwards
• RAM: 4GB onwards
• SSD/HDD: 256GB/ 512 GB onwards
• Ubuntu or (any Linux variant), iOS, Windows (less
preferred)
• Desktop, laptop, mobile phone, tablet, iPads and
RaspberryPi kit
Instructor,GroupIDs and TAs
• Instructor and Course coordinator: Dr. Yamuna
Prasad (yamuna.prasad@iitjammu.ac.in,
+91-9354941189)

• TAs: Ankur [Intellipaat] - +91-9519412446


• Mr. Mahendra Kumar Gurve [IIT Jammu]
(2021rcs2011@iitjammu.ac.in) +91-9871160843
Code of Conduct
• Encourage students to form study groups
• Students may do discussions
• Each student must write down the program & solutions
independently
• Each student must understand the solution well enough in order
to reconstruct it by him/herself.
• Each student should submit his/her own code and mention
anyone he/she collaborated with.
• Violation to copy, refer to, or look at written or code solutions
from a previous year, including but not limited to: official solutions
from a previous year, solutions posted online, and solutions you
or someone else may have written up in a previous year.
• In case of Conduct Violation by any means will be highly
penalised.
Introduction:Learning Process

AATCCGCTAG…………

Artificial
Intelligence

Input: Output:
Machine Learning
Artifical Data Action
Artificial
Sensors Intelligence Decision
PathLabs Text

Deep
Learning

Real Data Trained Model Prediction

Img Ref: web


Introduction:Learning Process
contd…
• Data Cleansing + Data Preparation + Data Analysis
• Data Cleansing: Noise Removal
• Data Preparation: Missing Value Curation, (Multi
Dimensional) Vector Space Representation,
Feature Identification, Feature Extraction, Feature
Vector Normalisation, Data Augmentation without
bias, Data Visualization
• Data Analysis: Statistical Inferencing, Machine
Learning Models and Other Complex Predictive
Models
Applications: Games
Applications: Smart City
Applications: Jeopardy Game
Applications: Robots in Service
Applications: Robots in
Medicine
Applications: Blockchain
Applications: Ethics??
Applications: Ethics?? Contd…
Motivation for ML (Multi-label
!24
classification)

Tag this image with the set of relevant labels from


{Male, Narendra Modi, Attractive, American President }
Motivation for ML (Ranking)
!25

Rank these people in increasing order of


attractiveness.
Motivation for ML (Social Media)
!26

User Location, Sentiment analysis,


Spam vs Non-spam
27
Spam ?
28
Spam ?
29
NLP Domain
30

Examples:
WEBSPAM, KDDB, NEWS20, REUTERS etc.
(Millions of dimensions)
NLP Domain: Translation
31

http://sampark.org.in/sampark/web/index.php/content
NLP Domain: Summarization
32

http://textsummarization.net/text-summarizer
NLP Domain: Summarization
33

http://text-processing.com/demo/sentiment/
NLP Domain: NER
34

http://cogcomp.cs.illinois.edu/page/demo_view/ner
NLP Domain: NER
35

http://cogcomp.cs.illinois.edu/page/demo_view/ner
NLP Domain: Q&A
36

http://start.csail.mit.edu/index.php
NLP Domain: Other Probs.
37

Co-referincing
News Article Title Generation
Code Categorization
Question Answering (Eliza)
Question Paper Analysis etc.
Vision Domain
38

Task: Classification Challenge: Redundant Features


Solution: Dimensionality Reduction
Examples: MNIST, USPS, Datasets for action recognition etc.
(Thousands to millions of dimensions)
Face Detection

Face interface Face database


Face detection
Face recognition
Output:
Mr. X
Face detection Face recognition
Mr. Y

!39 face interface v.2a


Face Detection

!40
Digit/Character Recognition
Vision Domain: Other
Applications
● Colorization of Black and White Images.
● Adding Sounds To Silent Movies.
● Automatic Machine Translation.
● Object Classification in Photographs.
● Automatic Handwriting Generation.
● Character Text Generation.
● Image Caption Generation.
● Automatic Game Playing.
Image Coloring
Image Coloring
Video Coloring

Colorizebot Neural Network Research (Deep Learning) - Coloring a B&W Video-clip Number 2.mp4
Sound Embedding to Salient
Movies

Visually-Indicated Sounds.mp4
Object Classification
Object detection + Caption
Hand Writing Generation

http://www.cs.toronto.edu/~graves/handwriting.html
Image Caption Generation
Art to Stylish Image
Game Playing

IBM's Watson Supercomputer Destroys Humans in Jeopardy - Engadget.mp4

DQN Breakout.mp4

Google DeepMind's Deep Q-learning playing Atari Breakout


Bioinformatics Domain (Gene
53
Expression Data)

Task: Classify novel samples into


known disease type (disease
diagnosis)
Challenge: Thousands of genes
(columns), few samples
Solution: Dimensionality Reduction

Examples: Colon, Lymphoma, Leukemia, RAOA, T2D, RAC etc.


54

You might also like