Scribd
Upload
164 views

Intro To ML

This document provides an overview of an online machine learning course taught by Professor Carl Gustaf Jansson. The 8-week course covers fundamental machine learning topics like supervised and unsupervised learning, different representation forms, inductive learning techniques, and neural networks. It is intended as a broad introduction to machine learning rather than focusing on specific algorithms, languages, or applications. The course also discusses relationships between machine learning, statistics, data science, cognitive science, and other relevant fields.

Uploaded by

Shruthin Gatty NMAMIT
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
164 views

Intro To ML

This document provides an overview of an online machine learning course taught by Professor Carl Gustaf Jansson. The 8-week course covers fundamental machine learning topics like supervised and unsupervised learning, different representation forms, inductive learning techniques, and neural networks. It is intended as a broad introduction to machine learning rather than focusing on specific algorithms, languages, or applications. The course also discusses relationships between machine learning, statistics, data science, cognitive science, and other relevant fields.

Uploaded by

Shruthin Gatty NMAMIT
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 18

NPTEL

Video Course on Machine Learning

Professor Carl Gustaf Jansson, KTH

Week 1 Introduction to the


Machine Learning course

Video 1.1 Introduction to the


Machine Learning course
Welcome to this course on Machine Learning

The purpose of this course is to give a broad introduction to Machine Learning as


being as subfield of Artificial Intelligence and Computer Science.

The course is NOT aimed to be a professional course in the sense of giving hands-on
knowledge on specific algorithms coded in specific languages applied to specific
problems. You can find many such alternative courses.

Machine Learning has imported much relevant knowledge from statistics and
probability theory but this is NOT a general course on these fields either.

This course will hopefully make you a better computer scientist who are able to
handle data analysis problems, but it will NOT make you a general statistician or data
scientist.
Artificial Intelligence

Machine Learning

Statistics Data Science

Big Data

Incremental Large Distributed Incremental


Sensor Data Databases User Data
This is NOT a course on Statistics
Statistics is an established discipline with a long history.

Statistics has traditionally dealt with:


- data collection
- data modelling
- data analysis
- data interpretation
- data presentation

The core theoretical part of statistics: Mathematical statistics is


based on Probability Theory, a branch of Mathematics.

Machine Learning is mostly dependent on inferential statistics which draws conclusions


on populations from studies of samples in contrast to descriptive statistics, which
primarily summarizes samples.
This is NOT a course on Data Science either
Data Science was launched as the new umbrella concept for Statistics
augmented with other Data analysis and Machine Learning techniques.

Data Science as for Statistics is assumed to cover:


- data collection/data capturing/data harvesting
- data modeling
- data maintenance
- data analysis/data processing
- visualization/presentation of data and decision making based on data.
The concept was coined at some lectures and conferences in 1996-1998 (Chikio
Hayashi and C.F. Jeff Wu) and has been further boosted in the last decade.
Publicity like the Harvard Business Review article in 2012 : "Data Scientist: The
Sexiest Job of the 21st Century“ has contributed.
This is NOT a course on Big Data either
In contrast to Data Science, the term Big Data primary refers to the storage, maintenance
and access to data, even if the border line with Data Science is somewhat blurred. The Big
Data area is based on more traditional areas such as very large databases, data warehousing
and distributed databases.

The area was born in the late 1990ies out of the fast data growth due to the explosive
developent of harvested sensor data as well as user data from the web and the parallel growth
of traditional databases in a variety of sectors.

When we speak about Big Data we talk about Terabytes (10^12) and already up to Zettabytes
(10^21), but the size is is not the only relevant parameter, the variety of data types, the quality
of data and the velocity of generation of data are also important.
Overview of Course

Week 1 Introduction to the Machine Learning course


Week 2 Characterization of Learning Problems
Week 3 Forms of Representation

Week 4 Inductive Learning based on Symbolic Representations


with Weak Theories
Week 5 Learning based on Symbolic Representations
with Strong Theories
Week 6 Machine Learning in Artificial Neural Networks

Week 7 Machine Learning in the Context of Cognitive Science


Week 8 Tools and Resources
Dependencies between weeks

Week 1

Week 2

Week 7 Week 3 Week 8

Week 4 Week 5 Week 6


Week 1 Machine Learning course Introduction

Lecture 1.1 Introduction to the Machine Learning course

Lecture 1.2 Foundation of Artificial Intelligence and Machine Learning


Lecture 1.3 Intelligent Autonomous Systems and Artificial Intelligence
Lecture 1.4 Applications of Machine Learning

Lecture 1.5 Tutorial for Week 1


Week 2 Characterization of Learning Problems

Lecture 2.1 Characterization of Learning Problems

Lecture 2.2 Objects, Categories and Features


Lecture 2.3 Roles of objects in learning processes
Lecture 2.4 Supervised contra Unsupervised learning
Lecture 2.5 Batch contra Incremental learning
Lecture 2.6 Generalization contra Instance based learning
Lecture 2.7 Weak or Strong Theories
Lecture 2.8 Learning Representations
Lecture 2.9 Reinforcement learning

Lecture 2.10 Tutorial for week 2


Week 3 Forms of Representation

Lecture 3.1 Forms of Representation

Lecture 3.2 Decision Trees


Lecture 3.3 Bayesian Networks
Lecture 3.4 Artificial Neural Networks
Lecture 3.5 Genetic Algorithms
Lecture 3.6 Semantic Networks and Production Rules
Lecture 3.7 Logical and functional programs

Lecture 3.8 Tutorial for week 3


Week 4 Inductive Learning based on Symbolic Representations
and Weak Theories

Lecture 4.1 Inductive Learning based on Symbolic Representations


and Weak Theories

Lecture 4.2 Generalization as Search


Lecture 4.3 Learning of Decision trees
Lecture 4.4 Bayesian Learning
Lecture 4.5 Instance-based learning
Lecture 4.6 Clustering Techniques

Lecture 4.7 Tutorial for Week 4


Week 5 Learning based on Symbolic Representations
with Strong Theories

Lecture 5.1 Learning based on Symbolic Representations


with Strong Theories

Lecture 5.2 Inductive Logic Programming


Lecture 5.3 Explanation-based learning
Lecture 5.4 Learning by Analogy
Lecture 5.5 Learning by Being Told
Lecture 5.6 Re-inforcement learning

Lecture 5.7 Tutorial for Week 5


Week 6 Machine Learning in Artificial Neural Networks

Lecture 6.1 Machine Learning in Artificial Neural Networks

Lecture 6.2 Perceptrons


Lecture 6.3 Backpropagation in Artificial Neural Networks
Lecture 6.4 Recurrent Networks and Convolution Networks
Lecture 6.5 Hopfield Networks and Boltzman Machines
Lecture 6.6 Deep Learning

Lecture 6.7 Tutorial for Week 6


Week 7 Machine Learning in the
Context of Cognitive Science

Lecture 7.1 Machine Learning in the Context of Cognitive Science

Lecture 7.2 Models in Cognitive Psychology


Lecture 7.3 Categorization
Lecture 7.4 Human Problem Solving and Human Expertise
Lecture 7.5 Models in Neuro Science
Lecture 7.6 Perception
Lecture 7.7 Theory of Evolution

Lecture 7.8 Tutorial for Week 7


Week 8 Tools and Resources

Lecture 8.1 Tools and Resources

Lecture 8.2 General Technical Computing Tools


Lecture 8.3 General Software support
Lecture 8.4 Deep Learning Support Platforms
Lecture 8.5 Large standarized openly available Datasets

Lecture 8.6 Tutorial for Week 8


General texts on Machine Learning

Machine Learning, Tom Mitchell, McGraw Hill, 1997.

https://www.cs.ubbcluj.ro/~gabis/ml/ml-books/McGrawHill%20-%20Machine%20Learning%20-Tom%20Mitchell.pdf

Introduction to Machine Learning, Ethem Alpaydin, MIT Press, 2014.

Deep Learning, Ian Goodfellow, Yoshua Bengio, Aaron Courville, MIT Press, 2016.
NPTEL

Video Course on Machine Learning

Professor Carl Gustaf Jansson, KTH

Thanks for your attention!

The next lecture 1.2 will be on the


topic:

Foundation of Artificial Intelligence


and Machine Learning

You might also like