INTRO TO Pembelajaran Mesin

D4 TI

MACHINE LEARNING IFY

IT Del 2019
WHAT IS MACHINE LEARNING?
A branch of artificial intelligence, concerned with the design and
development of algorithms that allow computers to evolve
behaviors based on empirical data.

As intelligence requires knowledge, it is necessary for the

computers to acquire knowledge.
LEARNING SYSTEM MODEL

Testing

Input Learning
Samples Method

System

Training
TRAINING AND TESTING

Data acquisition Practical usage

Universal set
(unobserved)

Training set Testing set

(observed) (unobserved)
 TRAINING AND TESTING
Training is the process of making the system able to learn.

No free lunch rule:

 Training set and testing set come from the same distribution
 Need to make some assumptions or bias
 TRAINING, DEVELOPMENT, AND TESTING
SET
Convention of data distribution (in the term of ML):
 Hold-out method: ratio (80%: 10%: 10%); (90%: 5%: 5%).
 (K-Fold) cross-validation: dividing dataset into k parts. Each k-th part will be
as train, development, or testing set.

What is the difference between leave-one-out cross-

validation and k-fold cross-validation?
Is it OK to implement hold-out approach if the sample
data is very small? Another methodology?
Wednesday, 18 September Machine TMP
2019 Learning
 PERFORMANCE
There are several factors affecting the performance:
 Types of training provided
 The form and extent of any initial background knowledge
 The type of feedback provided
 The learning algorithms used

Two important factors:

 Modeling
 Optimization
 ALGORITHMS
The success of machine learning system also depends on the
algorithms.

The algorithms control the search to find and build the knowledge
structures.

The learning algorithms should extract useful information from

training examples.
 ALGORITHMS
Supervised learning
 Prediction
 Classification (discrete labels), Regression (real values)

Unsupervised learning
 Clustering
 Probability distribution estimation
 Finding association (in features)
 Dimension reduction

Semi-supervised learning
Reinforcement learning
 Decision making (robot, chess machine)
ALGORITHMS

Supervised learning Unsupervised learning

Semi-supervised learning 10
 MACHINE LEARNING STRUCTURE
Supervised learning Unsupervised learning
WHY MACHINE LEARNING?
No human experts
 industrial/manufacturing control
 mass spectrometer analysis, drug design, astronomic discovery
Black-box human expertise
 face/handwriting/speech recognition
 driving a car, flying a plane
Rapidly changing phenomena
 credit scoring, financial modeling
 diagnosis, fraud detection
Need for customization/personalization
 personalized news reader
 movie/book recommendation
RELATED FIELDS
data
mining control theory

statistics
decision theory
information theory machine
learning
cognitive science
databases
psychological models
evolutionary neuroscience
models

Machine learning is primarily concerned with the accuracy and

effectiveness of the computer system.
 ARCHITECTURE OF A LEARNING SYSTEM
feedback performance standard
critic percepts

ENVIRONMENT
changes
learning performance
element element actions
knowledge
learning goals

problem
generator
MACHINE LEARNING ARCHITECHTURE
SUMBER
Yi-Fan Chang, Prof. J. J. Ding, 2011, An Overview of Machine Learning
SRI International, Machine Learning: An Overview
TMP, 2018, Introduction to Machine Learning