Machine Learning in AI
Download as PPTX, PDF1 like85 views
The document introduces the fundamentals of machine learning, defining it as a process of improving performance through experiences or concept building. It discusses key elements like memorization, approximation, and adaptation, illustrating how these components relate to learning from data. Examples include predicting stock values and understanding relationships between variables like height and weight, highlighting the need for effective algorithm design in machine learning.
Machine Learning in AI
- 2. MODULE 1 BASICS OF MACHINE LEARNING
- 3. SECTION 1.1 ELEMENTS OF MACHINE LEARNING
- 4. • In general learning means doing better on experiences or building concepts - • so does Machine Learning What is Machine Learning Elements of Machine Learning 3
- 5. • Learning is used when o Relationships in data are not clear • Stock value forecast o Humans are unable to explain their expertise • speech recognition, image recognition o Human expertise does not exist • navigating on Mars o Solution changes in time • routing on a computer network Why Machine Learning Elements of Machine Learning 4
- 6. • Memorization is the very basic form of the learning. • Who will be quick to recall an information - man or machine? Ability to Recall ~ Learning Elements of Machine Learning 5
- 7. • What is relationship between height and weight of a person? Approximation ~ Learning Elements of Machine Learning 6 Height(cm) Weight(kg) 125 32 130 35 118 28 140 42 170 85 190 86 155 65 143 60 177 80 162 158 W ~ H - 100 0 20 40 60 80 100 120 140 160 180 0 50 100 150 200 Weight(kg) An approximation model : W ~ H – 100 A generic approximation: W ~ αH + β
- 8. • What distinguishes a better approximation from a good approximation? Optimization ~ Learning Elements of Machine Learning 7
- 9. • Adaptation often known as learnability is probably most distinguishing component of intelligent behavior Adaptation ~ Learning Elements of Machine Learning 8 Old Data (Area A) Height(cm) Weight(kg) 125 32 130 35 118 28 140 42 170 85 190 86 155 65 143 60 177 80 162 158 Old model: W ~ H - 100 More Data (Area B) Height(cm) Weight(kg) 124 36 131 38 170 95 140 45 130 40 190 86 155 65 175 85 162 158 New model on this data ? New model on overall data ?
- 10. 4 Elements of Machine Learning Elements of Machine Learning 9 Data Objective Algorithm Design Machine Learning
Editor's Notes
- #2: Cover Option2
- #3: Welcome to module-1 – basics of machine learning
- #4: Seciton 1.1: elements of machine learning
- #5: Let us make a simple definition – for now In general learning means doing better on experiences or building concepts - so does Machine Learning
- #6: Lec 1.1 (Basics of ML)
- #8: Machine learning as approximation ~ finding alpha and beta that fits best on given data.
- #9: Often there are is no perfect approximation or there are more than candidate approximations, which one should we select.
- #11: Lecture 1.2 (Types of ML)
- #12: Cover Option2