CSCI 3346, Fall 2018
Prof. Alvarez

Exam 1 Topics 
(also see PS 1-4, their solutions, and the relevant sections of the textbook)

The KDD process
	Preprocessing, data mining, postprocessing

Supervised and unsupervised data mining tasks
        Classification, regression, clustering

Evaluation of data mining techniques
        Accuracy, error rate
	Training performance vs. generalization performance
	Cross-validation

Data preprocessing
        Sampling, outlier removal, Mahalanobis distance
        Discretization, class entropy
        Attribute selection, feature extraction

Exploratory data analysis
        Summary statistics, correlation, covariance

Model complexity vs data complexity
	Underfitting and overfitting
	Bias-variance tradeoff
	MDL principle

Decision tree classification
        Tree induction algorithm
	Early stopping, pruning, statistical error bounds

Rule-based classification
	Coverage and accuracy
	Sequential covering approach
	
Similarity and instance-based prediction
        Similarity metrics
        k-NN classifier / regressor