Question? Leave a message!




Data Mining: Data

Data Mining: Data
Data Mining: Data © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›What is Data  Collection of data objects and Attributes their attributes Tid Refund Marital Taxable  An attribute is a property or Cheat Status Income characteristic of an object 1 Yes Single 125K No – Examples: eye color of a 2 No Married 100K No person, temperature, etc. 3 No Single 70K No – Attribute is also known as 4 Yes Married 120K No variable, field, characteristic, 5 No Divorced 95K Yes or feature Objects 6 No Married 60K No  A collection of attributes 7 Yes Divorced 220K No describe an object 8 No Single 85K Yes – Object is also known as 9 No Married 75K No record, point, case, sample, 10 No Single 90K Yes entity, or instance 10 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Attribute Values  Attribute values are numbers or symbols assigned to an attribute  Distinction between attributes and attribute values – Same attribute can be mapped to different attribute values  Example: height can be measured in feet or meters – Different attributes can be mapped to the same set of values  Example: Attribute values for ID and age are integers  But properties of attribute values can be different – ID has no limit but age has a maximum and minimum value © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Measurement of Length  The way you measure an attribute is somewhat may not match the attributes properties. A 5 1 B 7 2 C 8 3 D 10 4 E 15 5 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Types of Attributes  There are different types of attributes – Nominal  Examples: ID numbers, eye color, zip codes – Ordinal  Examples: rankings (e.g., taste of potato chips on a scale from 110), grades, height in tall, medium, short – Interval  Examples: calendar dates, temperatures in Celsius or Fahrenheit. – Ratio  Examples: temperature in Kelvin, length, time, counts © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Properties of Attribute Values  The type of an attribute depends on which of the following properties it possesses: – Distinctness: =  – Order: – Addition: + – Multiplication: / – Nominal attribute: distinctness – Ordinal attribute: distinctness order – Interval attribute: distinctness, order addition – Ratio attribute: all 4 properties © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Attribute Description Examples Operations Type Nominal The values of a nominal attribute are zip codes, employee mode, entropy, just different names, i.e., nominal ID numbers, eye color, contingency 2 attributes provide only enough sex: male, female correlation,  test information to distinguish one object from another. (=, ) Ordinal The values of an ordinal attribute hardness of minerals, median, percentiles, provide enough information to order good, better, best, rank correlation, objects. (, ) grades, street numbers run tests, sign tests Interval For interval attributes, the calendar dates, mean, standard differences between values are temperature in Celsius deviation, Pearson's meaningful, i.e., a unit of or Fahrenheit correlation, t and F measurement exists. tests (+, ) Ratio For ratio variables, both differences temperature in Kelvin, geometric mean, and ratios are meaningful. (, /) monetary quantities, harmonic mean, counts, age, mass, percent variation length, electrical currentAttribute Transformation Comments Level Nominal Any permutation of values If all employee ID numbers were reassigned, would it make any difference Ordinal An order preserving change of An attribute encompassing values, i.e., the notion of good, better newvalue = f(oldvalue) best can be represented where f is a monotonic function. equally well by the values 1, 2, 3 or by 0.5, 1, 10. Interval newvalue =a oldvalue + b Thus, the Fahrenheit and where a and b are constants Celsius temperature scales differ in terms of where their zero value is and the size of a unit (degree). Ratio newvalue = a oldvalue Length can be measured in meters or feet.Discrete and Continuous Attributes  Discrete Attribute – Has only a finite or countably infinite set of values – Examples: zip codes, counts, or the set of words in a collection of documents – Often represented as integer variables. – Note: binary attributes are a special case of discrete attributes  Continuous Attribute – Has real numbers as attribute values – Examples: temperature, height, or weight. – Practically, real values can only be measured and represented using a finite number of digits. – Continuous attributes are typically represented as floatingpoint variables. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Types of data sets  Record – Data Matrix – Document Data – Transaction Data  Graph – World Wide Web – Molecular Structures  Ordered – Spatial Data – Temporal Data – Sequential Data – Genetic Sequence Data © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Important Characteristics of Structured Data – Dimensionality  Curse of Dimensionality – Sparsity  Only presence counts – Resolution  Patterns depend on the scale © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Record Data  Data that consists of a collection of records, each of which consists of a fixed set of attributes Tid Refund Marital Taxable Cheat Status Income 1 Yes Single 125K No 2 No Married 100K No 3 No Single 70K No 4 Yes Married 120K No 5 No Divorced 95K Yes 6 No Married 60K No 7 Yes Divorced 220K No 8 No Single 85K Yes 9 No Married 75K No 10 No Single 90K Yes 10 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Data Matrix  If data objects have the same fixed set of numeric attributes, then the data objects can be thought of as points in a multidimensional space, where each dimension represents a distinct attribute  Such data set can be represented by an m by n matrix, where there are m rows, one for each object, and n columns, one for each attribute P Pr ro ojje ec ct tiio on n P Pr ro ojje ec ct tiio on n D Diis st ta an nc ce e L Lo oa ad d T Th hiic ck kn ne es ss s o of f x x L Lo oa ad d o of f y y llo oa ad d 1 10 0..2 23 3 5 5..2 27 7 1 15 5..2 22 2 2 2..7 7 1 1..2 2 1 12 2..6 65 5 6 6..2 25 5 1 16 6..2 22 2 2 2..2 2 1 1..1 1 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›season timeout lost wi n game score ball pla y coach team Document Data  Each document becomes a `term' vector, – each term is a component (attribute) of the vector, – the value of each component is the number of times the corresponding term occurs in the document. Document 1 3 0 5 0 2 6 0 2 0 2 Document 2 0 7 0 2 1 0 0 3 0 0 Document 3 0 1 0 0 1 2 2 0 3 0 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Transaction Data  A special type of record data, where – each record (transaction) involves a set of items. – For example, consider a grocery store. The set of products purchased by a customer during one shopping trip constitute a transaction, while the individual products that were purchased are the items. TID Items 1 Bread, Coke, Milk 2 Beer, Bread 3 Beer, Coke, Diaper, Milk 4 Beer, Bread, Diaper, Milk 5 Coke, Diaper, Milk © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Graph Data  Examples: Generic graph and HTML Links a href="papers/papers.htmlbbbb" Data Mining /a li a href="papers/papers.htmlaaaa" 2 Graph Partitioning /a li a href="papers/papers.htmlaaaa" 1 5 Parallel Solution of Sparse Linear System of Equations /a li 2 a href="papers/papers.htmlffff" NBody Computation and Dense Linear System Solvers 5 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Chemical Data  Benzene Molecule: C H 6 6 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Ordered Data  Sequences of transactions Items/Events An element of the sequence © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Ordered Data  Genomic sequence data GGTTCCGCCTTCAGCCCCGCGCC CGCAGGGCCCGCCCCGCGCCGTC GAGAAGGGCCCGCCTGGCGGGCG GGGGGAGGCGGGGCCGCCCGAGC CCAACCGAGTCCGACCAGGTGCC CCCTCTGCTCGGCCTAGACCTGA GCTCATTAGGCGGCAGCGGACAG GCCAAGTAGAACACGCGAAGCGC TGGGCTGCCTGCTGCGACCAGGG © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Ordered Data  SpatioTemporal Data Average Monthly Temperature of land and ocean © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Data Quality  What kinds of data quality problems  How can we detect problems with the data  What can we do about these problems  Examples of data quality problems: – Noise and outliers – missing values – duplicate data © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Noise  Noise refers to modification of original values – Examples: distortion of a person’s voice when talking on a poor phone and “snow” on television screen Two Sine Waves Two Sine Waves + Noise © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Outliers  Outliers are data objects with characteristics that are considerably different than most of the other data objects in the data set © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Missing Values  Reasons for missing values – Information is not collected (e.g., people decline to give their age and weight) – Attributes may not be applicable to all cases (e.g., annual income is not applicable to children)  Handling missing values – Eliminate Data Objects – Estimate Missing Values – Ignore the Missing Value During Analysis – Replace with all possible values (weighted by their probabilities) © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Duplicate Data  Data set may include data objects that are duplicates, or almost duplicates of one another – Major issue when merging data from heterogeous sources  Examples: – Same person with multiple email addresses  Data cleaning – Process of dealing with duplicate data issues © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Data Preprocessing  Aggregation  Sampling  Dimensionality Reduction  Feature subset selection  Feature creation  Discretization and Binarization  Attribute Transformation © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Aggregation  Combining two or more attributes (or objects) into a single attribute (or object)  Purpose – Data reduction  Reduce the number of attributes or objects – Change of scale  Cities aggregated into regions, states, countries, etc – More “stable” data  Aggregated data tends to have less variability © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Aggregation Variation of Precipitation in Australia Standard Deviation of Average Standard Deviation of Average Monthly Precipitation Yearly Precipitation © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Sampling  Sampling is the main technique employed for data selection. – It is often used for both the preliminary investigation of the data and the final data analysis.  Statisticians sample because obtaining the entire set of data of interest is too expensive or time consuming.  Sampling is used in data mining because processing the entire set of data of interest is too expensive or time consuming. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Sampling …  The key principle for effective sampling is the following: – using a sample will work almost as well as using the entire data sets, if the sample is representative – A sample is representative if it has approximately the same property (of interest) as the original set of data © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Types of Sampling  Simple Random Sampling – There is an equal probability of selecting any particular item  Sampling without replacement – As each item is selected, it is removed from the population  Sampling with replacement – Objects are not removed from the population as they are selected for the sample.  In sampling with replacement, the same object can be picked up more than once  Stratified sampling – Split the data into several partitions; then draw random samples from each partition © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Sample Size 8000 points 2000 Points 500 Points © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Sample Size  What sample size is necessary to get at least one object from each of 10 groups. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Curse of Dimensionality  When dimensionality increases, data becomes increasingly sparse in the space that it occupies  Definitions of density and distance between points, which is critical for clustering and outlier detection, become less meaningful • Randomly generate 500 points • Compute difference between max and min distance between any pair of points © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Dimensionality Reduction  Purpose: – Avoid curse of dimensionality – Reduce amount of time and memory required by data mining algorithms – Allow data to be more easily visualized – May help to eliminate irrelevant features or reduce noise  Techniques – Principle Component Analysis – Singular Value Decomposition – Others: supervised and nonlinear techniques © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Dimensionality Reduction: PCA  Goal is to find a projection that captures the largest amount of variation in data x 2 e x 1 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Dimensionality Reduction: PCA  Find the eigenvectors of the covariance matrix  The eigenvectors define the new space x 2 e x 1 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Dimensionality Reduction: ISOMAP By: Tenenbaum, de Silva, Langford (2000)  Construct a neighbourhood graph  For each pair of points in the graph, compute the shortest path distances – geodesic distances © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Dimensionality Reduction: PCA D D D D D D im im im im im im ens ens ens ens ens ens ions ions ions ions ions ions = = = = = = 206 160 120 80 10 40 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Feature Subset Selection  Another way to reduce dimensionality of data  Redundant features – duplicate much or all of the information contained in one or more other attributes – Example: purchase price of a product and the amount of sales tax paid  Irrelevant features – contain no information that is useful for the data mining task at hand – Example: students' ID is often irrelevant to the task of predicting students' GPA © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Feature Subset Selection  Techniques: – Bruteforce approch: Try all possible feature subsets as input to data mining algorithm – Embedded approaches:  Feature selection occurs naturally as part of the data mining algorithm – Filter approaches:  Features are selected before data mining algorithm is run – Wrapper approaches:  Use the data mining algorithm as a black box to find best subset of attributes © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Feature Creation  Create new attributes that can capture the important information in a data set much more efficiently than the original attributes  Three general methodologies: – Feature Extraction  domainspecific – Mapping Data to New Space – Feature Construction  combining features © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Mapping Data to a New Space  Fourier transform  Wavelet transform Two Sine Waves Two Sine Waves + Noise Frequency © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Discretization Using Class Labels  Entropy based approach 3 categories for both x and y 5 categories for both x and y © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Discretization Without Using Class Labels Data Equal interval width Equal frequency Kmeans © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Attribute Transformation  A function that maps the entire set of values of a given attribute to a new set of replacement values such that each old value can be identified with one of the new values k x – Simple functions: x , log(x), e , x – Standardization and Normalization © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Similarity and Dissimilarity  Similarity – Numerical measure of how alike two data objects are. – Is higher when objects are more alike. – Often falls in the range 0,1  Dissimilarity – Numerical measure of how different are two data objects – Lower when objects are more alike – Minimum dissimilarity is often 0 – Upper limit varies  Proximity refers to a similarity or dissimilarity © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Similarity/Dissimilarity for Simple Attributes p and q are the attribute values for two data objects. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Euclidean Distance  Euclidean Distance n 2 dist( p q ) k k k1 Where n is the number of dimensions (attributes) and p and q k k th are, respectively, the k attributes (components) or data objects p and q.  Standardization is necessary, if scales differ. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Euclidean Distance 3 point x y p1 2 p1 0 2 p3 p4 p2 2 0 1 p3 3 1 p2 p4 5 1 0 0 1 2 3 4 5 6 p1 p2 p3 p4 p1 0 2.828 3.162 5.099 p2 2.828 0 1.414 3.162 p3 3.162 1.414 0 2 p4 5.099 3.162 2 0 Distance Matrix © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Minkowski Distance  Minkowski Distance is a generalization of Euclidean Distance 1 n r r dist ( p q )  k k k1 Where r is a parameter, n is the number of dimensions (attributes) and p and q are, respectively, the kth attributes k k (components) or data objects p and q. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Minkowski Distance: Examples  r = 1. City block (Manhattan, taxicab, L norm) distance. 1 – A common example of this is the Hamming distance, which is just the number of bits that are different between two binary vectors  r = 2. Euclidean distance  r. “supremum” (L norm, L norm) distance. max  – This is the maximum difference between any component of the vectors  Do not confuse r with n, i.e., all these distances are defined for all numbers of dimensions. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Minkowski Distance L1 p1 p2 p3 p4 p1 0 4 4 6 p2 4 0 2 4 p3 4 2 0 2 p4 6 4 2 0 point x y L2 p1 p2 p3 p4 p1 0 2 p1 0 2.828 3.162 5.099 p2 2 0 p2 2.828 0 1.414 3.162 p3 3 1 p3 3.162 1.414 0 2 p4 5 1 p4 5.099 3.162 2 0 L p1 p2 p3 p4  p1 0 2 3 5 p2 2 0 1 3 p3 3 1 0 2 p4 5 3 2 0 Distance Matrix © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Mahalanobis Distance 1 T mahalanobis( p,q) ( p q) ( p q)  is the covariance matrix of the input data X n 1  (X X )(X X ) j k j,k ij ik n1 i1 For red points, the Euclidean distance is 14.7, Mahalanobis distance is 6. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Mahalanobis Distance Covariance Matrix: 0.3 0.2    0.2 0.3  C A: (0.5, 0.5) B B: (0, 1) A C: (1.5, 1.5) Mahal(A,B) = 5 Mahal(A,C) = 4 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Common Properties of a Distance  Distances, such as the Euclidean distance, have some well known properties. 1. d(p, q)  0 for all p and q and d(p, q) = 0 only if p = q. (Positive definiteness) 2. d(p, q) = d(q, p) for all p and q. (Symmetry) 3. d(p, r)  d(p, q) + d(q, r) for all points p, q, and r. (Triangle Inequality) where d(p, q) is the distance (dissimilarity) between points (data objects), p and q.  A distance that satisfies these properties is a metric © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Common Properties of a Similarity  Similarities, also have some well known properties. 1. s(p, q) = 1 (or maximum similarity) only if p = q. 2. s(p, q) = s(q, p) for all p and q. (Symmetry) where s(p, q) is the similarity between points (data objects), p and q. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Similarity Between Binary Vectors  Common situation is that objects, p and q, have only binary attributes  Compute similarities using the following quantities M = the number of attributes where p was 0 and q was 1 01 M = the number of attributes where p was 1 and q was 0 10 M = the number of attributes where p was 0 and q was 0 00 M = the number of attributes where p was 1 and q was 1 11  Simple Matching and Jaccard Coefficients SMC = number of matches / number of attributes = (M + M ) / (M + M + M + M ) 11 00 01 10 11 00 J = number of 11 matches / number of notbothzero attributes values = (M ) / (M + M + M ) 11 01 10 11 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›SMC versus Jaccard: Example p = 1 0 0 0 0 0 0 0 0 0 q = 0 0 0 0 0 0 1 0 0 1 M = 2 (the number of attributes where p was 0 and q was 1) 01 M = 1 (the number of attributes where p was 1 and q was 0) 10 M = 7 (the number of attributes where p was 0 and q was 0) 00 M = 0 (the number of attributes where p was 1 and q was 1) 11 SMC = (M + M )/(M + M + M + M ) = (0+7) / (2+1+0+7) = 0.7 11 00 01 10 11 00 J = (M ) / (M + M + M ) = 0 / (2 + 1 + 0) = 0 11 01 10 11 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Cosine Similarity  If d and d are two document vectors, then 1 2 cos( d , d ) = (d d ) / d d , 1 2 1 2 1 2 where indicates vector dot product and d is the length of vector d.  Example: d = 3 2 0 5 0 0 0 2 0 0 1 d = 1 0 0 0 0 0 0 1 0 2 2 d d = 31 + 20 + 00 + 50 + 00 + 00 + 00 + 21 + 00 + 02 = 5 1 2 0.5 0.5 d = (33+22+00+55+00+00+00+22+00+00) = (42) = 6.481 1 0.5 0.5 d = (11+00+00+00+00+00+00+11+00+22) = (6) = 2.245 2 cos( d , d ) = .3150 1 2 © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Extended Jaccard Coefficient (Tanimoto)  Variation of Jaccard for continuous or count attributes – Reduces to Jaccard for binary attributes © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Correlation  Correlation measures the linear relationship between objects  To compute correlation, we standardize data objects, p and q, and then take their dot product  p ( p mean( p)) / std ( p) k k  q (q mean(q)) / std (q) k k  correlation( p,q) p q © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Visually Evaluating Correlation Scatter plots showing the similarity from –1 to 1. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›General Approach for Combining Similarities  Sometimes attributes are of many different types, but an overall similarity is needed. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Using Weights to Combine Similarities  May not want to treat all attributes the same. – Use weights w which are between 0 and 1 and sum k to 1. © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Density  Densitybased clustering require a notion of density  Examples: – Euclidean density  Euclidean density = number of points per unit volume – Probability density – Graphbased density © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Euclidean Density – Cellbased  Simplest approach is to divide region into a number of rectangular cells of equal volume and define density as of points the cell contains © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›Euclidean Density – Centerbased  Euclidean density is the number of points within a specified radius of the point © Tan,Steinbach, Kumar Introduction to Data Mining 4/18/2004 ‹›