Introduction to Machine Learning Yazd University Electrical and

  • Slides: 16
Download presentation
Introduction to Machine Learning Yazd University, Electrical and Computer Engineering Department Course Title: Advanced

Introduction to Machine Learning Yazd University, Electrical and Computer Engineering Department Course Title: Advanced Software Engineering By: Mohammad Ali Zare Chahooki 1

Why “Learn” ? �Machine learning is programming computers to … optimize a performance criterion

Why “Learn” ? �Machine learning is programming computers to … optimize a performance criterion using … example data or past experience. �There is no need to “learn” to calculate payroll �Learning is used when: �Human expertise does not exist (navigating on 2 Mars), �Humans are unable to explain their expertise (speech recognition) �Solution changes in time (routing on a computer network) �Solution needs to be adapted to particular cases (user biometrics)

In “Learning” … �Learning general models from a data of particular examples �Data is

In “Learning” … �Learning general models from a data of particular examples �Data is cheap and abundant (data warehouses); knowledge is expensive and scarce. �Example in retail … People who bought “Da Vinci Code” also bought “The Five People You Meet in Heaven” (www. amazon. com) �Build a model that is a good and useful approximation to the data. 3

What is Machine Learning? �Optimize a performance criterion using example data or past experience.

What is Machine Learning? �Optimize a performance criterion using example data or past experience. �Role of Statistics: Inference from a sample �Role of Computer science: Efficient algorithms to �Solve the optimization problem �Representing and evaluating the model for inference 4

Applications �Learning Associations �Supervised Learning �Classification �Regression �Unsupervised Learning �Reinforcement Learning 5

Applications �Learning Associations �Supervised Learning �Classification �Regression �Unsupervised Learning �Reinforcement Learning 5

Learning Associations �Basket analysis: P (Y | X ) probability that somebody who buys

Learning Associations �Basket analysis: P (Y | X ) probability that somebody who buys X also buys Y where X and Y are products/services. Example: P ( chips | delester) = 0. 7 6

Classification �Example: Credit scoring �Differentiating between low-risk and high-risk customers from their income and

Classification �Example: Credit scoring �Differentiating between low-risk and high-risk customers from their income and savings 7 Discriminant: IF income > θ 1 AND savings > θ 2 THEN low-risk ELSE high-risk

Classification: Applications �Face recognition: Pose, lighting, occlusion (glasses, beard), make-up, hair style �Character recognition:

Classification: Applications �Face recognition: Pose, lighting, occlusion (glasses, beard), make-up, hair style �Character recognition: Different handwriting styles. �Speech recognition: Temporal dependency. �Use of a dictionary or the syntax of the language. �Sensor fusion: Combine multiple modalities; eg, visual (lip image) and acoustic for speech �Medical diagnosis: From symptoms to illnesses �. . . 8

Face Recognition Training examples of a person Test images AT&T Laboratories, Cambridge UK http:

Face Recognition Training examples of a person Test images AT&T Laboratories, Cambridge UK http: //www. uk. research. att. com/facedatabase. html 9

Regression �Example: Price of a used car �x : car attributes y : price

Regression �Example: Price of a used car �x : car attributes y : price y = g (x | θ ) g ( ) model, θ parameters 10 y = wx+w 0

Regression Applications �Navigating a car: Angle of the steering wheel (CMU Nav. Lab) �Kinematics

Regression Applications �Navigating a car: Angle of the steering wheel (CMU Nav. Lab) �Kinematics of a robot arm (x, y) α 2 α 1 11 α 1= g 1(x, y) α 2= g 2(x, y)

Supervised Learning: Uses �Prediction of future cases: Use the rule to predict the output

Supervised Learning: Uses �Prediction of future cases: Use the rule to predict the output for future inputs �Knowledge extraction: The rule is easy to understand �Compression: The rule is simpler than the data it explains �Outlier detection: Exceptions that are not covered by the rule, e. g. , fraud 12

Unsupervised Learning �In supervised learning, the aim is to … learn a mapping from

Unsupervised Learning �In supervised learning, the aim is to … learn a mapping from the input to an output 13 whose … correct values are provided by a supervisor. �In unsupervised learning, there is no such supervisor and … we only have input data. The aim is to find the regularities in the input. �One method is clustering where the aim is to find clusters or groupings of input. �Other methods like feature reduction and finding association rules

Reinforcement Learning �In some applications, the output of the system is a sequence of

Reinforcement Learning �In some applications, the output of the system is a sequence of actions. �In such a case, a single action is not important; what is important is the policy that is the sequence of correct actions to reach the goal. �There is no such thing as the best action in any intermediate state; … an action is good if it is part of a good policy. 14

Reinforcement Learning �In such a case, the machine learning program should be able to

Reinforcement Learning �In such a case, the machine learning program should be able to … assess the goodness of policies and … learn from past good action sequences to be able to generate a policy. �Such learning methods are called 15 reinforcement learning algorithms. �A good example is game playing where … a single move by itself is not that important; it is the sequence of right moves that is

Reference �E. Alpaydın, "Introduction to Machine Learning 2 nd edition Ed. “, MIT Press,

Reference �E. Alpaydın, "Introduction to Machine Learning 2 nd edition Ed. “, MIT Press, (2010) �E. Alpaydın, "Introduction to Machine Learning 3 nd edition Ed. “, MIT Press, (2014) 16

YAZD Yazd a veces transliterada como Yezd esYAZD Yazd a veces transliterada como Yezd es
YAZD Yazd parfois translittr comme Yezd est luneYAZD Yazd parfois translittr comme Yezd est lune
Machine Learning Applications in Software Engineering Yazd UniversityMachine Learning Applications in Software Engineering Yazd University
The Project Management Discipline Yazd University Electrical andThe Project Management Discipline Yazd University Electrical and
The Rational Unified Process RUP Yazd University ElectricalThe Rational Unified Process RUP Yazd University Electrical
Machine Learning Machine Learning Machine learningstatistical learning artificialMachine Learning Machine Learning Machine learningstatistical learning artificial
What is Machine Learning Machine learning defined MachineWhat is Machine Learning Machine learning defined Machine
Machine Learning Spark machine learning Spark Mllib MachineMachine Learning Spark machine learning Spark Mllib Machine
What is Machine Learning Machine learning defined MachineWhat is Machine Learning Machine learning defined Machine
Pneumonia Dr Samet M Apr 2011 Yazd UniversityPneumonia Dr Samet M Apr 2011 Yazd University
Retinitis Pigmentosa MR Manaviat Yazd university of MedicalRetinitis Pigmentosa MR Manaviat Yazd university of Medical
Derivative based optimization Vali Derhami Yazd University ComputerDerivative based optimization Vali Derhami Yazd University Computer
MicroprocessorBased Systems Dr Sadr Yazd University Instruction SetMicroprocessorBased Systems Dr Sadr Yazd University Instruction Set
MicroprocessorBased Systems Sadrossadat Yazd University Interrupts Exceptions TrapsMicroprocessorBased Systems Sadrossadat Yazd University Interrupts Exceptions Traps
MicroprocessorBased Systems Dr Sadr Yazd University Clocks CountersMicroprocessorBased Systems Dr Sadr Yazd University Clocks Counters