Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
22.02.17 Characterisedata use Big_University_DB mine characteristics as "Science_Students" in relevance to name,gender,major,birth_date,residence,pho ne#,gpa from student where statusin graduate DataMiningMTAT.03.183 Descriptiveanalysis,preprocessing, visualisation... JaakVilo 2017Spring JaakViloandotherauthors Knowledge Discovery (KDD) Process n Data in the real world is dirty n incomplete: lacking attribute values, lacking certain attributes of interest, or containing only aggregate data n Task-relevant Data n Selection Data Cleaning n Databases n DataMining:ConceptsandTechniques February 22, 2017 n n n n n n n n n Different data sources Functional dependency violation (e.g., modify some linked data) n Duplicate records also need data cleaning Data Mining: Concepts and Techniques Quality decisions must be based on quality data n Faulty data collection instruments Human or computer error at data entry Errors in data transmission February 22, 2017 4 No quality data, no quality mining results! n Inconsistent data may come from n n n “Not applicable” data value when collected Different considerations between the time when the data was collected and when it is analyzed. Human/hardware/software problems Noisy data (incorrect values) may come from n Data Mining: Concepts and Techniques Why Is Data Preprocessing Important? Incomplete data may come from n e.g., Age=“42” Birthday=“03/07/1997” e.g., Was rating “1,2,3”, now rating “A, B, C” e.g., discrepancy between duplicate records 3 Why Is Data Dirty? n e.g., Salary=“-10” inconsistent: containing discrepancies in codes or names n Data Integration e.g., occupation=“ ” noisy: containing errors or outliers n n Jiawei Han, Micheline Kamber, and Jian Pei 2 Why Data Preprocessing? • This is a view from typical database systems and data Pattern Evaluation warehousing communities • Data mining plays an essential role in the knowledge discovery Data Mining process Data Warehouse UT:DataMining2009 5 e.g., duplicate or missing data may cause incorrect or even misleading statistics. Data warehouse needs consistent integration of quality data Data extraction, cleaning, and transformation comprises the majority of the work of building a data warehouse Garbage in, garbage out February 22, 2017 Data Mining: Concepts and Techniques 6 1 22.02.17 High quality data requirements n n n n n n n n n Multi-Dimensional Measure of Data Quality High-quality data needs to pass a set of quality criteria. Those include: Accuracy: an aggregated value over the criteria of integrity, consistency, and density Integrity: an aggregated value over the criteria of completeness and validity Completeness: achieved by correcting data containing anomalies Validity: approximated by the amount of data satisfying integrity constraints Consistency: concerns contradictions and syntactical anomalies Uniformity: directly related to irregularities and in compliance with the set 'unit of measure' Density: the quotient of missing values in the data and the number of total values ought to be known http://en.wikipedia.org/wiki/Data_cleansing February 22, 2017 n Data Mining: Concepts and Techniques n n 7 A well-accepted multidimensional view: n Accuracy n Completeness n Consistency n Timeliness n Believability n Value added n Interpretability n Accessibility Broad categories: n Intrinsic, contextual, representational, and accessibility February 22, 2017 Major Tasks in Data Preprocessing n 8 Forms of Data Preprocessing Data cleaning n Fill in missing values, smooth noisy data, identify or remove outliers, and resolve inconsistencies n Data integration n Data transformation n Data reduction n n n n Data Mining: Concepts and Techniques Integration of multiple databases, data cubes, or files Normalization and aggregation Obtains reduced representation in volume but produces the same or similar analytical results Data discretization n Part of data reduction but with particular importance, especially for numerical data February 22, 2017 Data Mining: Concepts and Techniques 9 February 22, 2017 Data Mining: Concepts and Techniques 10 Chapter 2: Data Preprocessing n Why preprocess the data? n Descriptive data summarization n Data cleaning n Data integration and transformation n Data reduction n Discretization and concept hierarchy generation n Summary • 0.480.030.060.050.430.190.160.350.250.07 0.290.140.960.020.110.220.800.050.540.36 0.230.280.020.100.480.310.360.210.330.45 0.640.040.480.560.160.580.330.110.420.06 0.000.230.240.000.540.020.260.200.180.01 0.170.170.040.970.250.040.340.010.500.15 0.430.050.500.160.520.820.230.090.020.21 0.130.170.330.260.000.330.570.430.090.43 0.240.080.080.540.080.020.020.010.350.62 0.100.030.140.780.300.070.080.480.570.30 JaakViloandotherauthors February 22, 2017 Data Mining: Concepts and Techniques UT:DataMining2009 12 11 2 22.02.17 1,20# 1,20# 1,00# 1,00# 0,80# 0,80# 0,60# 0,60# Series1# Series1# 0,40# 0,40# 0,20# 0,20# 0,00# 0# 20# 40# 60# 80# 100# 120# 0,00# 0# 20# 40# 60# 80# 100# 120# JaakViloandotherauthors 1,20# 15 1,00# UT:DataMining2009 14 0,80# JaakViloandotherauthors UT:DataMining2009 0,60# JaakViloandotherauthors 0,40# 13 0,20# UT:DataMining2009 0,00# JaakViloandotherauthors 0# 20# 40# 60# 80# 100# 120# n n n To better understand the data: central tendency, variation and spread n median, max, min, quantiles, outliers, variance, etc. n n Weighted arithmetic mean: n Trimmed mean: chopping extreme values x= Median: A holistic measure n n Boxplot or quantile analysis on sorted intervals x N åw x i =1 n i i åw i =1 i Middle value if odd number of values, or average of the middle two Estimated by interpolation (for grouped data): median = L1 + ( Mode n Value that occurs most frequently in the data n Folding measures into numerical dimensions n Unimodal, bimodal, trimodal n Boxplot or quantile analysis on the transformed cube n Empirical formula: Dispersion analysis on computed measures Data Mining: Concepts and Techniques µ=å n values otherwise Data dispersion: analyzed with multiple granularities of precision February 22, 2017 1 n å xi n i =1 Mean (algebraic measure) (sample vs. population): x = n Numerical dimensions correspond to sorted intervals n n n Data dispersion characteristics n 16 Measuring the Central Tendency Motivation n Series1# Mining Data Descriptive Characteristics UT:DataMining2009 17 February 22, 2017 n / 2 - (å f )l f median )c mean - mode = 3 ´ (mean - median ) Data Mining: Concepts and Techniques 18 3 22.02.17 • • HistogramsandProbabilityDensityFunctions ProbabilityDensityFunctions • FrequencyHistograms – • Totalareaundercurveintegratesto1 – – Y-axisiscounts Simpleinterpretation – Can'tbedirectlyrelatedtoprobabilitiesordensityfunctions RelativeFrequencyHistograms – – Dividecountsbytotalnumberofobservations Y-axisisrelativefrequency – – Canbeinterpretedasprobabilitiesforeachrange Can'tbedirectlyrelatedtodensityfunction • • Barheightssumto1butwon'tintegrateto1unlessbarwidth=1 DensityHistograms – – – – Dividecountsby(totalnumberofobservationsXbarwidth) Y-axisisdensityvalues BarheightXbarwidthgivesprobabilityforeachrange Canbedirectlyrelatedtodensityfunction • Barareassumto1 http://www.geog.ucsb.edu/~joel/g210_w07/lecture_notes/lect04/oh07_04_1.html JaakViloandotherauthors UT:DataMining2009 19 JaakViloandotherauthors UT:DataMining2009 20 JaakViloandotherauthors UT:DataMining2009 21 JaakViloandotherauthors UT:DataMining2009 22 histograms • equalsub-intervals,knownas`bins‘ • breakpoints • binwidth The data are (the log of) wing spans of aircraft built in from 1956 - 1984. JaakViloandotherauthors UT:DataMining2009 23 JaakViloandotherauthors UT:DataMining2009 24 4 22.02.17 JaakViloandotherauthors UT:DataMining2009 25 2-dimensionaldata:dot-plot JaakViloandotherauthors UT:DataMining2009 26 Histogramvskerneldensity • propertiesofhistogramswiththesetwoexamples: – theyarenotsmooth – dependonendpointsofbins – dependonwidthofbins • Wecanalleviatethefirsttwoproblemsbyusing kerneldensityestimators. • Toremovethedependenceontheendpointsofthe bins,wecentreeachoftheblocksateachdatapoint ratherthanfixingtheendpointsoftheblocks. DataMining:ConceptsandTechniques February 22, 2017 27 JaakViloandotherauthors UT:DataMining2009 28 • Blocks- itisstilldiscontinuousaswehave usedadiscontinuouskernelasourbuilding block • Ifweuseasmoothkernelforourbuilding block,thenwewillhaveasmoothdensity estimate. block of width 1 and height 1/12 (the dotted boxes) as they are 12 data points, and then add them up JaakViloandotherauthors UT:DataMining2009 29 JaakViloandotherauthors UT:DataMining2009 30 5 22.02.17 • It'simportanttochoosethemostappropriate bandwidthasavaluethatistoosmallortoolargeis notuseful. • Ifweuseanormal(Gaussian)kernelwithbandwidth orstandarddeviationof0.1(whichhasarea1/12 undertheeachcurve)thenthekerneldensity estimateissaidtoundersmoothedasthebandwidth istoosmallinthefigurebelow. • Itappearsthatthereare4modesinthisdensitysomeofthesearesurelyartificesofthedata. JaakViloandotherauthors UT:DataMining2009 31 JaakViloandotherauthors UT:DataMining2009 32 • Chooseoptimalbandwith –Methodstoestimateit • AMISE=AsymptoticMeanIntegratedSquaredError • optimalbandwidth=argminAMISE JaakViloandotherauthors UT:DataMining2009 33 JaakViloandotherauthors UT:DataMining2009 34 JaakViloandotherauthors UT:DataMining2009 36 • Theoptimalvalueofthebandwidthforourdatasetis about0.25. • Fromtheoptimallysmoothedkerneldensity estimate,therearetwomodes.Asthesearethelog ofaircraftwingspan,itmeansthattherewerea groupofsmaller,lighterplanesbuilt,andtheseare clusteredaround2.5(whichisabout12m). • Whereasthelargerplanes,maybeusingjetengines astheseusedonacommercialscalefromaboutthe 1960s,aregroupedaround3.5(about33m). JaakViloandotherauthors UT:DataMining2009 35 6 22.02.17 KernelDensityestimation- links • Thepropertiesofkerneldensityestimators are,ascomparedtohistograms: • Wikipedia: http://en.wikipedia.org/wiki/Kernel_density_estimation • RicardoGutierrez-Osuna – smooth – noendpoints – dependonbandwidth http://research.cs.tamu.edu/prism/lectures/pr/pr_l7.pdf • TutorialandJavaappletfortesting: – http://parallel.vub.ac.be/research/causalModels/tutorial/kde.html • SimpleinteractivewebdemoatU.Tartu: – https://courses.cs.ut.ee/demos/kernel-density-estimation/ – http://kde.tume-maailm.pri.ee/ – JaakViloandotherauthors UT:DataMining2009 37 JaakViloandotherauthors UT:DataMining2009 38 R– example(dueK.Tretjakov) d=c(1,2,2,2,2,1,2,2,2,3,2,3,4,5,4,3,2,3,4,4,5,6,7); kernelsmooth<- function(data,sigma,x){ result=0; for(dindata){ result=result+exp(-(x-d)^2/2/sigma^2); } result/sqrt(2*pi)/sigma; } x=seq(min(d),max(d),by=0.1); y=sapply(x,function(x){kernelsmooth(d,1,x)}); hist(d); lines(x,y); JaakViloandotherauthors UT:DataMining2009 39 JaakViloandotherauthors UT:DataMining2009 40 41 JaakViloandotherauthors UT:DataMining2009 42 http://parallel.vub.ac.be/research/causalModels/tutorial/kde.html JaakViloandotherauthors UT:DataMining2009 7 22.02.17 http://kde.tume-maailm.pri.ee/ JaakViloandotherauthors UT:DataMining2009 43 http://kde.tume-maailm.pri.ee/ JaakViloandotherauthors JaakViloandotherauthors UT:DataMining2009 44 UT:DataMining2009 46 http://kde.tume-maailm.pri.ee/ UT:DataMining2009 45 JaakViloandotherauthors Aggregation,analysisand visualizationofgeodata • http://sightsmap.com • Severallargecrowd-sourceddatasets: – ThewholePanoramiophotobankusedbyGooglemaps – ThewholeWikipedia,geotagsandwikipediaarticlelogs – Foursquare – ...More • Aggregatedata,calculatepopularitesofplaces,calculatetypetags • Translateandcategorizetitlesanddescriptionstogettypes • Improveaggregationalgorithmsbylearning • Visualizeheatmaps,typetags,aggregatedsources By: Tanel Tammet By: Tanel Tammet 8 22.02.17 By: Tanel Tammet By: Tanel Tammet By: Tanel Tammet By: Tanel Tammet • http://176.32.89.45/~hideaki/res/kernel.html JaakViloandotherauthors UT:DataMining2009 53 JaakViloandotherauthors UT:DataMining2009 54 9 22.02.17 http://jmlr.csail.mit.edu/proceedings/papers/v2/kontkanen07a/kontkanen07a.pdf JaakViloandotherauthors UT:DataMining2009 55 JaakViloandotherauthors UT:DataMining2009 56 MorelinksonRandkerneldensity • http://en.wikipedia.org/wiki/Kernel_density_estimation • http://sekhon.berkeley.edu/stats/html/density.html • http://stat.ethz.ch/R-manual/Rpatched/library/stats/html/density.html • http://www.google.com/search?q=kernel+density+estimation+R JaakViloandotherauthors UT:DataMining2009 57 JaakViloandotherauthors UT:DataMining2009 58 JaakViloandotherauthors UT:DataMining2009 59 JaakViloandotherauthors UT:DataMining2009 60 10 22.02.17 JaakViloandotherauthors UT:DataMining2009 61 JaakViloandotherauthors UT:DataMining2009 62 Symmetric vs. Skewed Data n Median, mean and mode of symmetric, positively and negatively skewed data Author: Gea Pajula JaakViloandotherauthors UT:DataMining2009 63 February 22, 2017 Quartiles, outliers and boxplots n Quartiles: Q1 (25th percentile), Q3 (75th percentile) n Inter-quartile range: IQR = Q3 – Q1 n Five number summary: min, Q1, M, Q3, max n Boxplot: ends of the box are the quartiles, median is marked, whiskers, and n Five-number summary of a distribution: n Boxplot Minimum, Q1, M, Q3, Maximum n plot outlier individually n n n Outlier: usually, a value higher/lower than 1.5 x IQR Variance and standard deviation (sample: s, population: σ) n Variance: (algebraic, scalable computation) 1 n 1 n 2 1 n 2 s = xi - (å xi ) ] å ( xi - x )2 = n - 1[å n - 1 i =1 n i =1 i =1 2 n 64 Boxplot Analysis Measuring the Dispersion of Data n Data Mining: Concepts and Techniques 1 s = N 2 n 1 ( xi - µ ) = å N i =1 2 n åx i =1 i 2 -µ 2 Data is represented with a box The ends of the box are at the first and third quartiles, i.e., the height of the box is IRQ n The median is marked by a line within the box n Whiskers: two lines outside the box extend to Minimum and Maximum Standard deviation s (or σ) is the square root of variance s2 (or σ2) February 22, 2017 Data Mining: Concepts and Techniques 65 February 22, 2017 Data Mining: Concepts and Techniques 66 11 22.02.17 BoxPlots • Tukey77:JohnW.Tukey,"ExploratoryData Analysis".Addison-Wesley,Reading,MA. 1977. • http://informationandvisualization.de/blog/box-plot • JaakViloandotherauthors UT:DataMining2009 67 JaakViloandotherauthors UT:DataMining2009 68 1.5 x IQR – Inter Quartile range JaakViloandotherauthors UT:DataMining2009 JaakViloandotherauthors UT:DataMining2009 69 JaakViloandotherauthors UT:DataMining2009 70 71 JaakViloandotherauthors UT:DataMining2009 72 Gea Pajula 12 22.02.17 Visualization of Data Dispersion: Boxplot Analysis February 22, 2017 Data Mining: Concepts and Techniques 73 February 22, 2017 Data Mining: Concepts and Techniques 74 Violin plot February 22, 2017 Data Mining: Concepts and Techniques 75 Violin plot – R Combining plots… vertical kernel density http://gallery.r-enthusiasts.com/graph/Violin_plot,43 76 77 13 22.02.17 Properties of Normal Distribution Curve n February 22, 2017 Data Mining: Concepts and Techniques 79 February 22, 2017 n February 22, 2017 Data Mining: Concepts and Techniques 81 Quantile-Quantile (q-q) Plots n n n The normal (distribution) curve n From μ–σ to μ+σ: contains about 68% of the measurements (μ: mean, σ: standard deviation) n From μ–2σ to μ+2σ: contains about 95% of it n From μ–3σ to μ+3σ: contains about 99.7% of it Data Mining: Concepts and Techniques 80 Example of plots-containing article: n http://www.kgs.ku.edu/Magellan/WaterLevels/ CD/Reports/OFR04_57/rep00.htm February 22, 2017 Data Mining: Concepts and Techniques 82 Cumulative Distribution Function (CDF) http://onlinestatbook.com/2/advanced_graphs/q-q_plots.html http://www.itl.nist.gov/div898/handbook/eda/section3/eda33o.htm A q-q plot is a plot of the quantiles of the first data set against the quantiles of the second data set. By a quantile, we mean the fraction (or percent) of points below the given value. That is, the 0.3 (or 30%) quantile is the point at which 30% percent of the data fall below and 70% fall above that value. February 22, 2017 Data Mining: Concepts and Techniques 83 February 22, 2017 Data Mining: Concepts and Techniques 84 14 22.02.17 n n n February 22, 2017 Data Mining: Concepts and Techniques 85 These 2 batches do not appear to have come from populations with a common distribution. The batch 1 values are significantly higher than the corresponding batch 2 values. The differences are increasing from values 525 to 625. Then the values for the 2 batches get closer again. February 22, 2017 Data Mining: Concepts and Techniques 86 The advantages of the q-q plot are: n n February 22, 2017 Data Mining: Concepts and Techniques 87 The sample sizes do not need to be equal. Many distributional aspects can be simultaneously tested. For example, shifts in location, shifts in scale, changes in symmetry, and the presence of outliers can all be detected from this plot. For example, if the two data sets come from populations whose distributions differ only by a shift in location, the points should lie along a straight line that is displaced either up or down from the 45-degree reference line. February 22, 2017 Data Mining: Concepts and Techniques 88 A Q–Q plot comparing the distributions of standardizeddaily maximum temperatures at 25 stations in the US state of Ohio in March and in July. The curved pattern suggests that the central quantiles are more closely spaced in July than in March, and that the March distribution is skewed to the right compared to the July distribution. The data cover the period 1893–2001. February 22, 2017 Data Mining: Concepts and Techniques 89 February 22, 2017 Data Mining: Concepts and Techniques 90 15 22.02.17 n Parametric modeling usually involves making assumptions about the shape of data, or the shape of residuals from a regression fit. Verifying such assumptions can take many forms, but an exploration of the shape using histograms and qq plots is very effective. The q-q plot does not have any design parameters such as the number of bins for a histogram. February 22, 2017 Data Mining: Concepts and Techniques 91 February 22, 2017 Kemmeren et.al. (Mol. Cell, 2002) Data Mining: Concepts and Techniques 92 Scatter plot Randomized expression data Yeast 2-hybrid studies n Known (literature) PPI n Provides a first look at bivariate data to see clusters of points, outliers, etc Each pair of values is treated as a pair of coordinates and plotted as points in the plane MPK1 YLR350w SNF4 YCL046W SNF7 YGR122W. February 22, 2017 Data Mining: Concepts and Techniques 94 Pearson Correlation February 22, 2017 Data Mining: Concepts and Techniques 95 96 16 22.02.17 Not Correlated Data February 22, 2017 Data Mining: Concepts and Techniques 97 Abalone data set – P. Danelson hw. 98 Bias vs Variance Several sets of (x, y) points, with the Pearson correlation coefficient of x and y for each set. Note that the correlation reflects the noisiness and direction of a linear relationship (top row), but not the slope of that relationship (middle), nor many aspects of nonlinear relationships (bottom). N.B.: the figure in the center has a slope of 0 but in that case the correlation coefficient is undefined because the variance of Y is zero. Scott Fortmann-Roe http://scott.fortmann-roe.com/docs/BiasVariance.html 99 Numerical summary? February 22, 2017 Data Mining: Concepts and Techniques 100 Anscombe’s quartet 101 February 22, 2017 Data Mining: Concepts and Techniques 102 17 22.02.17 February 22, 2017 Data Mining: Concepts and Techniques 103 February 22, 2017 Data Mining: Concepts and Techniques 104 February 22, 2017 Data Mining: Concepts and Techniques 105 February 22, 2017 Data Mining: Concepts and Techniques 106 February 22, 2017 Data Mining: Concepts and Techniques 107 February 22, 2017 Data Mining: Concepts and Techniques 108 18 22.02.17 Loess Curve Adds a smooth curve to a scatter plot in order to provide better perception of the pattern of dependence Loess curve is fitted by setting two parameters: a smoothing parameter, and the degree of the polynomials that are fitted by the regression n n February 22, 2017 Data Mining: Concepts and Techniques 109 February 22, 2017 n Histogram: (shown before) Boxplot: (covered before) Quantile plot: each value xi is paired with fi indicating that approximately 100 fi % of data are £ xi n Quantile-quantile (q-q) plot: graphs the quantiles of one univariant distribution against the corresponding quantiles of another n Scatter plot: each pair of values is a pair of coordinates and plotted as points in the plane n Loess (local regression) curve: add a smooth curve to a scatter plot to provide better perception of the pattern of dependence n February 22, 2017 Data Mining: Concepts and Techniques 111 n Why preprocess the data? n Descriptive data summarization n Data cleaning n Data integration and transformation n Data reduction n Discretization and concept hierarchy generation n Summary February 22, 2017 n Importance n “Data cleaning is one of the three biggest problems in data warehousing”—Ralph Kimball n “Data cleaning is the number one problem in data warehousing”—DCI survey n n Fill in missing values n Identify outliers and smooth out noisy data n Correct inconsistent data n Resolve redundancy caused by data integration February 22, 2017 Data Mining: Concepts and Techniques 112 Data is not always available n Data cleaning tasks n Data Mining: Concepts and Techniques Missing Data Data Cleaning n 110 Chapter 2: Data Preprocessing Graphic Displays of Basic Statistical Descriptions n Data Mining: Concepts and Techniques Missing data may be due to n equipment malfunction n inconsistent with other recorded data and thus deleted n data not entered due to misunderstanding n n n 113 E.g., many tuples have no recorded value for several attributes, such as customer income in sales data certain data may not be considered important at the time of entry not register history or changes of the data Missing data may need to be inferred. February 22, 2017 Data Mining: Concepts and Techniques 114 19 22.02.17 How to Handle Missing Data? n Ignore the tuple: usually done when class label is missing (assuming the tasks in classification—not effective when the percentage of missing values per attribute varies considerably. n Fill in the missing value manually: tedious + infeasible? n Fill in it automatically with n a global constant : e.g., “unknown”, a new class?! n the attribute mean n the attribute mean for all samples belonging to the same class: smarter n the most probable value: inference-based such as Bayesian formula or decision tree February 22, 2017 Data Mining: Concepts and Techniques 115 February 22, 2017 Data Mining: Concepts and Techniques K-NN impute 116 Noisy Data n n K nearest neighbours imputation n Noise: random error or variance in a measured variable Incorrect attribute values may due to faulty data collection instruments data entry problems n data transmission problems n technology limitation n inconsistency in naming convention Other data problems which requires data cleaning n duplicate records n n Find K neighbours on available data points n Estimate the missing value n n (Hastie, Tibshirani, Troyanskaya, … Stanford 19992001) n n n February 22, 2017 Data Mining: Concepts and Techniques 117 February 22, 2017 How to Handle Noisy Data? n Binning n first sort data and partition into (equal-frequency) bins n then one can smooth by bin means, smooth by bin median, smooth by bin boundaries, etc. Regression n smooth by fitting the data into regression functions Clustering n detect and remove outliers Combined computer and human inspection n detect suspicious values and check by human (e.g., deal with possible outliers) February 22, 2017 Data Mining: Concepts and Techniques 118 Equal-width (distance) partitioning n Divides the range into N intervals of equal size: uniform grid n if A and B are the lowest and highest values of the attribute, the width of intervals will be: W = (B –A)/N. n n Data Mining: Concepts and Techniques Simple Discretization Methods: Binning n n incomplete data inconsistent data n n The most straightforward, but outliers may dominate presentation n Skewed data is not handled well Equal-depth (frequency) partitioning n Divides the range into N intervals, each containing approximately same number of samples 119 n Good data scaling n Managing categorical attributes can be tricky February 22, 2017 Data Mining: Concepts and Techniques 120 20 22.02.17 Binning Methods for Data Smoothing Regression Sorted data for price (in dollars): 4, 8, 9, 15, 21, 21, 24, 25, 26, 28, 29, 34 * Partition into equal-frequency (equi-depth) bins: - Bin 1: 4, 8, 9, 15 - Bin 2: 21, 21, 24, 25 - Bin 3: 26, 28, 29, 34 * Smoothing by bin means: - Bin 1: 9, 9, 9, 9 - Bin 2: 23, 23, 23, 23 - Bin 3: 29, 29, 29, 29 * Smoothing by bin boundaries: - Bin 1: 4, 4, 4, 15 - Bin 2: 21, 21, 25, 25 - Bin 3: 26, 26, 26, 34 y q February 22, 2017 Data Mining: Concepts and Techniques Y1 X1 121 February 22, 2017 Cluster Analysis G3 G2 n n February 22, 2017 Data Mining: Concepts and Techniques 123 February 22, 2017 n Data cleaning n Data integration and transformation n Data reduction n Discretization and concept hierarchy generation n Summary February 22, 2017 n Data Mining: Concepts and Techniques 122 Data Mining: Concepts and Techniques 124 Data Integration n Why preprocess the data? Data Mining: Concepts and Techniques Data discrepancy detection n Use metadata (e.g., domain, range, dependency, distribution) n Check field overloading n Check uniqueness rule, consecutive rule and null rule n Use commercial tools n Data scrubbing: use simple domain knowledge (e.g., postal code, spell-check) to detect errors and make corrections n Data auditing: by analyzing data to discover rules and relationship to detect violators (e.g., correlation and clustering to find outliers) Data migration and integration n Data migration tools: allow transformations to be specified n ETL (Extraction/Transformation/Loading) tools: allow users to specify transformations through a graphical user interface Integration of the two processes n Iterative and interactive (e.g., Potter’s Wheels) Chapter 2: Data Preprocessing n x Data Cleaning as a Process n G1 y=x+1 Y1’ n n 125 Data integration: n Combines data from multiple sources into a coherent store Schema integration: e.g., A.cust-id º B.cust-# n Integrate metadata from different sources Entity identification problem: n Identify real world entities from multiple data sources, e.g., Bill Clinton = William Clinton Detecting and resolving data value conflicts n For the same real world entity, attribute values from different sources are different n Possible reasons: different representations, different scales, e.g., metric vs. British units February 22, 2017 Data Mining: Concepts and Techniques 126 21 22.02.17 Handling Redundancy in Data Integration n Redundant data occur often when integration of multiple databases n n n n Object identification: The same attribute or object may have different names in different databases Derivable data: One attribute may be a “derived” attribute in another table, e.g., annual revenue Redundant attributes may be able to be detected by correlation analysis Careful integration of the data from multiple sources may help reduce/avoid redundancies and inconsistencies and improve mining speed and quality February 22, 2017 Data Mining: Concepts and Techniques 127 February 22, 2017 Data Mining: Concepts and Techniques 128 February 22, 2017 Data Mining: Concepts and Techniques 129 February 22, 2017 Data Mining: Concepts and Techniques 130 Normalisation n February 22, 2017 Data Mining: Concepts and Techniques 131 Making data comparable… February 22, 2017 Data Mining: Concepts and Techniques 132 22 22.02.17 Elements of microarray statistics 0,8% Reference 0,6% Test 0,4% 0,2% M = log2R – log2G = log2(R/G) A = 1/2 (log2R + log2G) Series1% Series2% Series3% Series4% 0% 0% 20% 40% 60% 80% 100% 120% !0,2% !0,4% !0,6% February 22, 2017 Data Mining: Concepts and Techniques 133 Normalisation can be used to transform data JaakViloandotherauthors UT:DataMining2009 135 Expression Profiler Data Reduction Strategies Chapter 2: Data Preprocessing n n Why preprocess the data? n Data cleaning n Data integration and transformation n Data reduction n Discretization and concept hierarchy generation n Summary n n February 22, 2017 Data Mining: Concepts and Techniques 136 137 Why data reduction? n A database/data warehouse may store terabytes of data n Complex data analysis/mining may take a very long time to run on the complete data set Data reduction n Obtain a reduced representation of the data set that is much smaller in volume but yet produce the same (or almost the same) analytical results Data reduction strategies n Data cube aggregation: n Dimensionality reduction — e.g., remove unimportant attributes n Data Compression n Numerosity reduction — e.g., fit data into models n Discretization and concept hierarchy generation February 22, 2017 Data Mining: Concepts and Techniques 138 23 22.02.17 Discretization Chapter 2: Data Preprocessing n Why preprocess the data? n Data cleaning n Data integration and transformation n Three types of attributes: n Nominal — values from an unordered set, e.g., color, profession n Ordinal — values from an ordered set, e.g., military or academic rank n n Data reduction n Discretization and concept hierarchy generation n n Summary February 22, 2017 Data Mining: Concepts and Techniques 139 Continuous — real numbers, e.g., integer or real numbers Discretization: n Divide the range of a continuous attribute into intervals n Some classification algorithms only accept categorical attributes. n Reduce data size by discretization n Prepare for further analysis February 22, 2017 Data Mining: Concepts and Techniques Discretization and Concept Hierarchy n E.g. Discretization n n Reduce the number of values for a given continuous attribute by n dividing the range of the attribute into intervals n n Interval labels can then be used to replace actual data values n Supervised vs. unsupervised n Split (top-down) vs. merge (bottom-up) n Discretization can be performed recursively on an attribute n n Concept hierarchy formation n n concepts (such as numeric values for age) by higher level concepts n (such as young, middle-aged, or senior) n Data Mining: Concepts and Techniques 141 0-18 y 19-65 y 66-105 1580 8394 2700 0-20 21-40 41-60 61-105 1780 4200 4767 2900 Vs n Recursively reduce the data by collecting and replacing low level February 22, 2017 140 February 22, 2017 Data Mining: Concepts and Techniques 142 Example of 3-4-5 Rule Segmentation by Natural Partitioning count n A simply 3-4-5 rule can be used to segment numeric data into relatively uniform, “natural” intervals. n If an interval covers 3, 6, 7 or 9 distinct values at the most significant digit, partition the range into 3 equi- Step 1: -$351 -$159 Min Low (i.e, 5%-tile) Step 2: msd=1,000 Step 3: (-$1,000 - 0) (-$400 - 0) significant digit, partition the range into 4 intervals n (-$400 -$300) If it covers 1, 5, or 10 distinct values at the most (-$300 -$200) significant digit, partition the range into 5 intervals February 22, 2017 Data Mining: Concepts and Techniques (-$200 -$100) 143 (-$100 0) $4,700 Max ($1,000 - $2,000) (0 -$ 1,000) (-$400 -$5,000) Step 4: If it covers 2, 4, or 8 distinct values at the most $1,838 High(i.e, 95%-0 tile) High=$2,000 (-$1,000 - $2,000) width intervals n profit Low=-$1,000 February 22, 2017 ($1,000 - $2, 000) (0 - $1,000) (0 $200) ($1,000 $1,200) ($200 $400) ($1,200 $1,400) ($600 $800) ($3,000 $4,000) ($1,400 $1,600) ($400 $600) ($800 $1,000) ($1,600 $1,800) ($2,000 - $5, 000) ($2,000 $3,000) ($1,800 $2,000) Data Mining: Concepts and Techniques ($4,000 $5,000) 144 24 22.02.17 Recursive … Example -351,976.00 .. 4,700,896.50 1.1. (-400,000 .. -300,000 ] 1.2. (-300,000 .. -200,000 ] 1.3. (-200,000 .. -100,000 ] 1.4. (-100,000 .. 0 ] MIN= -351,976.00 MAX=4,700,896.50 LOW = 5th percentile -159,876 HIGH = 95th percentile 1,838,761 msd = 1,000,000 (most significant digit) LOW = -1,000,000 (round down) HIGH = 2,000,000 (round up) 2.1. (0 .. 200,000 ] 2.2. (200,000 .. 400,000 ] 2.3. (400,000 .. 600,000 ] 2.4. (600,000 .. 800,000 ] 2.5. (800,000 .. 1,000,000 ] 3 value ranges 1. (-1,000,000 .. 0] 2. (0 .. 1,000,000] 3. (1,000,000 .. 2,000,000] 3.1. (1,000,000 .. 1,200,000 ] 3.2. (1,200,000 .. 1,400,000 ] 3.3. (1,400,000 .. 1,600,000 ] 3.4. (1,600,000 .. 1,800,000 ] 3.5. (1,800,000 .. 2,000,000 ] Adjust with real MIN and MAX 1. 2. 3. 4. (-400,000 .. 0] (0 .. 1,000,000] (1,000,000 .. 2,000,000] (2,000,000 .. 5,000,000] February 22, 2017 Data Mining: Concepts and Techniques 4.1. (2,000,000 .. 3,000,000 ] 4.2. (3,000,000 .. 4,000,000 ] 4.3. (4,000,000 .. 5,000,000 ] JaakViloandotherauthors 145 Concept Hierarchy Generation for Categorical Data • Specification of a partial/total ordering of attributes explicitly at the schema level by users or experts n – street < city < state < country Some hierarchies can be automatically generated based on the analysis of the number of distinct values per attribute in the data set n The attribute with the most distinct values is placed at the lowest level of the hierarchy n Exceptions, e.g., weekday, month, quarter, year – {Urbana, Champaign, Chicago} < Illinois 15 distinct values country • Specification of only a partial set of attributes – E.g., only street < city, not others • Automatic generation of hierarchies (or attribute levels) by the analysis of the number of distinct values – E.g., for a set of attributes: {street, city, state, country} DataMining:ConceptsandTechniques province_or_ state 365 distinct values city 3567 distinct values 674,339 distinct values street 147 February 22, 2017 Chapter 2: Data Preprocessing n Why preprocess the data? n Data cleaning n Data integration and transformation n Data reduction n Data Mining: Concepts and Techniques n n n Data preparation or preprocessing is a big issue for both data warehousing and data mining Discriptive data summarization is need for quality data preprocessing Data preparation includes n Data cleaning and data integration Discretization and concept hierarchy n Data reduction and feature selection generation n Discretization Summary February 22, 2017 Data Mining: Concepts and Techniques 148 Summary n n 146 Automatic Concept Hierarchy Generation • Specification of a hierarchy for a set of values by explicit data grouping February 22, 2017 UT:DataMining2009 149 A lot a methods have been developed but data preprocessing still an active area of research February 22, 2017 Data Mining: Concepts and Techniques 150 25 22.02.17 References n D. P. Ballou and G. K. Tayi. Enhancing data quality in data warehouse environments. Communications of ACM, 42:73-78, 1999 n T. Dasu and T. Johnson. Exploratory Data Mining and Data Cleaning. John Wiley & Sons, 2003 n T. Dasu, T. Johnson, S. Muthukrishnan, V. Shkapenyuk. Mining Database Structure; Or, How to Build a Data Quality Browser. SIGMOD’02. n H.V. Jagadish et al., Special Issue on Data Reduction Techniques. Bulletin of the Technical Committee on Data Engineering, 20(4), December 1997 n D. Pyle. Data Preparation for Data Mining. Morgan Kaufmann, 1999 n E. Rahm and H. H. Do. Data Cleaning: Problems and Current Approaches. IEEE Bulletin of the Technical Committee on Data Engineering. Vol.23, No.4 n V. Raman and J. Hellerstein. Potters Wheel: An Interactive Framework for Data Cleaning and Transformation, VLDB’2001 n T. Redman. Data Quality: Management and Technology. Bantam Books, 1992 n Y. Wand and R. Wang. Anchoring data quality dimensions ontological foundations. Communications of ACM, 39:86-95, 1996 n R. Wang, V. Storey, and C. Firth. A framework for analysis of data quality research. IEEE Trans. Knowledge and Data Engineering, 7:623-640, 1995 February 22, 2017 Data Mining: Concepts and Techniques 151 26