DOC PREVIEW
GSU CSC 2010 - Chapter 6 final

This preview shows page 1-2-20-21 out of 21 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 21 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 21 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 21 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 21 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 21 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

OutlineChapter 6. Classification and PredictionLazy vs. Eager LearningLazy Learner: Instance-Based MethodsThe k-Nearest Neighbor AlgorithmSlide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 12Fuzzy Set ApproachesSlide 14Slide 15Slide 16Slide 17Fuzzy Set ApplicationsSlide 19Classifier Accuracy MeasuresSlide 21OutlineK-Nearest Neighbor algorithmFuzzy Set theoryClassifier Accuracy MeasuresChapter 6. Classification and PredictionEager Learners: when given a set of training tuples, will construct a generalization model before receiving new tuples to classifyClassification by decision tree inductionRule-based classificationClassification by back propagationSupport Vector Machines (SVM)Associative classificationLazy vs. Eager LearningLazy vs. eager learningLazy learning (e.g., instance-based learning): Simply stores training data (or only minor processing) and waits until it is given a test tupleEager learning (the above discussed methods): Given a set of training set, constructs a classification model before receiving new (e.g., test) data to classifyLazy: less time in training but more time in predictingLazy Learner: Instance-Based MethodsTypical approachesk-nearest neighbor approachInstances represented as points in a Euclidean space.The k-Nearest Neighbor AlgorithmAll instances correspond to points in the n-D spaceThe nearest neighbor are defined in terms of Euclidean distance, dist(X1, X2)Target function could be discrete- or real- valuedFor discrete-valued, k-NN returns the most common value among the k training examples nearest to xq . _+_xq+_ _+__+The k-Nearest Neighbor Algorithmk-NN for real-valued prediction for a given unknown tupleReturns the mean values of the k nearest neighborsDistance-weighted nearest neighbor algorithmWeight the contribution of each of the k neighbors according to their distance to the query xqGive greater weight to closer neighborsRobust to noisy data by averaging k-nearest neighborsThe k-Nearest Neighbor AlgorithmHow can I determine the value of k, the number of neighbors?In general, the larger the number of training tuples is, the larger the value of k is Nearest-neighbor classifiers can be extremely slow when classifying test tuples O(n)By simple presorting and arranging the stored tuples into search tree, the number of comparisons can be reduced to O(logN)The k-Nearest Neighbor AlgorithmExample:K=5OutlineK-Nearest Neighbor algorithmFuzzy Set theoryClassifier Accuracy MeasuresFuzzy Set ApproachesRule-based systems for classification have the disadvantage that they involve sharp cutoffs for continuous attributesFor example:IF (years_employed>2) AND (income>50K)THEN credit_card=approvedWhat if a customer has 10 years employed and income is 49K?Fuzzy Set ApproachesInstead, we can discretize income into categories such as {low,medium,high}, and then apply fuzzy logic to allow “fuzzy” threshold for each categoryFuzzy Set ApproachesFuzzy theory is also known as possibility theory, it was proposed by Lotif Zadeh in 1965Unlike the notion of traditional “crisp” sets where an element either belongs to a set S, in fuzzy theory, elements can belong to more than one fuzzy setFuzzy Set ApproachesFor example, the income value $49K belongs to both the medium and high fuzzy sets:Mmedium($49K)=0.15 andMhigh($49K)=0.96Fuzzy Set ApproachesAnother example for temperatureFuzzy Set Applicationshttp://www.dementia.org/~julied/logic/applications.htmlOutlineK-Nearest Neighbor algorithmFuzzy Set theoryClassifier Accuracy MeasuresClassifier Accuracy Measuresclasses (Real) buy computer = yes(Real) buy computer = nototal(Predict) buy computer = yes6954 412 7366(Predict) buy computer = no46 2588 2634total 7000 (Buy Computer)3000 (Does not buy Computer)10000Classifier Accuracy MeasuresAlternative accuracy measures (e.g., for cancer diagnosis)sensitivity = t-pos/pos specificity = t-neg/neg precision = t-pos/(t-pos + f-pos)accuracy


View Full Document

GSU CSC 2010 - Chapter 6 final

Download Chapter 6 final
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Chapter 6 final and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Chapter 6 final 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?