Chapter 1

edu.omkar@gmail.
com
Chapter 1
1. Definition of learning systems: A computer program is said to learn from experience
E with respect to some class of tasks T and performance measure P, if its
performance at tasks in T, as measured by P, improves with experience E.
2. For well-defined(well-posed) Learning Problem, three features are important:
a. the class of tasks (recognizing, classifying, driving, i.e any activity)
b. the measure of performance to be improved (completeness, correcting,
wins, etc)
c. the source of experience (database, commands, practice etc)
3. Designing a Learning System:
a. Choosing the Training Experience: i.e choose the type of training experience
from which our system will learn.
i. Problem Faced: Credit Assignment
ii. Key Attributes:
1. Whether the training experience provides direct or indirect
feedback regarding the choices made by the performance
system.
2. The degree to which the learner controls the sequence of
training examples.
3. How well it represents the distribution of examples over which the
final system performance P must be measured
b. Choosing the Target Function: i.e determine exactly what type of knowledge
will be learned and how this will be used by the performance program
i. The process of learning the target function is often called function
approximation.
ii. we often expect learning algorithms to acquire only some
approximation to the target function V’ rather than ideal target
function V.
c. Choosing a Representation for the Target Function: i.e to describe the
function V’.
Example: Partial design of a checkers learning program:
• Task T: playing checkers
• Performance measure P: percent of games won in the world tournament
• Training experience E: games played against itself
• Target function: V: Board + R
• Target function representation:
The first three items above correspond to the specification of the learning task,
whereas the final two items constitute design choices for the implementation of the
learning program. Notice the net effect of this set of design choices is to reduce the
problem of learning a checkers strategy to the problem of learning values for the
coefficients wo through w6 in the target function representation.
d. Choosing a Function Approximation Algorithm: In order to learn the target

function V’ we require a set of training examples, each describing a specific
board state b and the training value Vtrain(b) for b. [Similar to Truth tables]
i. Ex:
(b, Vtrain(b))
ii. Steps:
[email protected]
1. ESTIMATING TRAINING VALUES:

(aage kya hoga, usko sochkar abhi ka decision lena (uss decision
ko rating dena))
2. ADJUSTING THE WEIGHTS:
a. BEST HYPOTHESIS: minimizing the squared error E between
the training values and the values predicted by the
hypothesis V’.
b. LMS TRAINGIN RULE (LEAST MEAN SQUARED)
i. Def: For each observed training example it adjusts
the weights a small amount in the direction that
reduces the error on this training example.
ii.
e. The Final Design: described by four distinct program modules
4. PERSPECTIVES IN MACHINE LEARNING

a. It involves searching a very large space of possible hypotheses to determine
one that best fits the observed data and any prior knowledge held by the
learner
[email protected]
5. Issues in Machine Learning

a. Algorithms: algorithms for learning general target functions, settings of algos,
best algos for various problems.
b. Training Data: How much? Hypothesis confidence, best strategy for choosing
a useful next training experience
c. Prior Knowledge
d. Function Approximation: 1 or more V’
e. Target Function
6. Classification, Regression, Supervised learning, Unsupervised Learning,
Reinforcement Learning – All in one brief (10 min read) at
https://towardsdatascience.com/types-of-machine-learning-algorithms-you-
should-know-953a08248861
Further Reading:
1. Classification in brief at https://www.simplilearn.com/classification-machine-
learning-tutorial
2. Regression in brief at https://www.simplilearn.com/regression-machine-learning-
tutorial
3. Supervised Learning at https://www.geeksforgeeks.org/ml-types-learning-
supervised-learning/?ref=rp
4. Unsupervised Learning at https://www.geeksforgeeks.org/ml-types-learning-part-
2/?ref=rp
5. Difference between Supervised and Unsupervised:
https://www.geeksforgeeks.org/difference-between-supervised-and-unsupervised-
learning/?ref=rp

Chapter 1

Uploaded by

Copyright:

Available Formats

Chapter 1

Uploaded by

Document Information

Original Description:

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

Chapter 1

Uploaded by

Copyright:

Available Formats

edu.omkar@gmail.

d. Choosing a Function Approximation Algorithm: In order to learn the target

1. ESTIMATING TRAINING VALUES:

e. The Final Design: described by four distinct program modules

4. PERSPECTIVES IN MACHINE LEARNING

5. Issues in Machine Learning

You might also like