# Learning from Observations - Donald Bren School of ...

Learning from Observations Chapter 18 Section 1 4 Outline Learning agents Inductive learning Nearest Neighbors

Learning agents Sometimes we want to invest time and effort to observe the feedback from our environment to our actions in order to improve these actions so we can more effectively optimize our utility in the future. Learning element

Design of a learning element is affected by Which components of the performance element are to be learned (e.g. learn to stop for traffic light) What feedback is available to learn these components (e.g. visual feedback form camera) What representation is used for the components (e.g. logic, probabilistic descriptions, attributes,...)

Type of feedback: Supervised learning: correct answers for each example (label). Unsupervised learning: correct answers not given. Reinforcement learning: occasional rewards Two Examples of Learning Object Categories. Here is your training set (2 classes):

Here is your test set: Does it belong to one of the above classes? Learning from 1 Example Copied from P. Perona talk slides.

S. Savarese, 2003 P. Buegel, 1562 Inductive learning Simplest form: learn a function from examples f is the target function

An example is a pair (x, f(x)) Problem: find a hypothesis h such that h f given a training set of examples Inductive learning method

Construct/adjust h to agree with f on training set (h is consistent if it agrees with f on all examples) E.g., curve fitting: Inductive learning method Construct/adjust h to agree with f on training set (h is consistent if it agrees with f on all examples) E.g., curve fitting:

Inductive learning method Inductive learning method Inductive learning method which curve is best?

Inductive learning method Ockhams razor: prefer the simplest hypothesis consistent with data Supervised Learning I Example: Imagine you want to classify

versus Data: 100 monkey images and 200 human images with labels what is what. {xi ,yi 0}, i 1,...,100 {x j ,yj 1}, j 1,...,200

where x represents the greyscale of the image pixels and y=0 means monkey while y=1 means human. Task: Here is a new image: monkey or human? 1 nearest neighbors

(your first ML algorithm!) Idea: 1. Find the picture in the database which is closest to your query image. 2. Check its label. 3. Declare the class of your query image to be the same as that of the closest picture. query

closest image 1NN Decision Surface decision curve Distance Metric

How do we measure what it means to be close? Depending on the problem we should choose an appropriate distance metric (or more generally, a (dis)similarity measure)

Hamming distance: D(xn , xm) | xn xm | Scaled Euclidean Distance:

D(xn , xm) (xn xm)T A (xn x m) {x discrete};

{x cont .}; -Demo: http://www.comp.lancs.ac.uk/~kristof/research/notes/nearb/cluster.html -Matlab Demo. Remarks on NN methods We only need to construct a classifier that works locally for each query. Hence: We dont need to construct a classifier everywhere in space.

Classifying is done at query time. This can be computationally taxing at a time where you might want to be fast. Memory inefficient. Curse of dimensionality: imagine many features are irrelevant / noisy distances are always large. Very flexible, not many prior assumptions. k-NN variants robust against bad examples.

Summary Learning agent = performance element + learning element For supervised learning, the aim is to find a simple hypothesis approximately consistent with training examples Decision tree learning + boosting Learning performance = prediction accuracy measured on test set

## Recently Viewed Presentations

• Ugly Bug Ball - sing along with Burl Ives Garden Spider Minibeast hunt Bugs! Bugs! Hey ho, how does it go? Mr Worm Minibeasts Everywhere Once a lonely caterpillar sat and cried To a sympathetic beetle by his side I've...
• Mirror-scenes between Richard and Anne (1.2) and Richard and Elizabeth (4.4): both full of stichomythic exchange, 'this keen encounter of our wits' (1.2.115). Does Richard or Elizabeth win the latter exchange? Metatheatre.
• Early Childhood Thought: Islands of Competence The Development of Children (5th ed.) Cole, Cole & Lightfoot Chapter 9 Early Childhood (age 2-6) Typical pattern of thinking in preschool years Mixture of sound logic and magical thinking Insight and ignorance The...
• New centers of civilization and political power arose. Much still remained in isolation. African Societies: Diversities and Similarities. Differences in geography, language, religion, politics, etc. meant that they never unified ... Nigeria- village of Nok- found objects demonstrating great ...
• David M, Schwartau I, Anand Pant H, Borde T. Emergency outpatient services in the city of Berlin: Factors for appropriate use and predictors for hospital admission. Eur J Emerg Med. 2006;13(6):352-7. 10.
• Agenda. Topic broken down into outline form, times for each subtopic. ... Your lesson plan. Agenda. Learning outcomes. Training methods. Evaluation strategy. Your process in creating the lesson plan. Challenges. Solutions. Something else? Lesson Planning for CSS: Summary.
• archived and made available in perpetuity. * These . goals apply both to peer-reviewed research publications, the data on which the results and conclusions of this research are based, and any software or code used in the course of the...
• Write an title that says what you are writing about (e.g. Jet travel: stop it or keep it?) Write an introduction that states what the discussion is about (There are many stories of sightings of a monster in Loch Ness....