Analytics in a Big Data World The Essential Guide to Data Science and its Applications by Bart Baesens (z-lib.org)

I N D E X ◂ 231 Scorecard, 161, 207 Stopword, 201 Application, 161 Supervised learning, 165 Behavioral, 163 Support vector machines, 58–64 Support vectors, 60, 62 Scoring, 136 Support, 87, 89, 94–95 Scree plot, 98–99 Suppression, 158 Search Engine Marketing Analytics, Survival analysis 193–194 evaluation, 117 Search engine optimization (SEO), 193 measurements, 106–109 Search term, 194 parametric, 111–114 Security, 151 semiparametric, 114–116 Segmentation, 32–33, 48, 95–96, 192 Survival function, 107 Self-organizing map (SOM), 100–102 baseline, 116 Senior management, 159 System stability index (SSI), 136, Sensitivity, 74 143 analysis, 92 Swing clients, 170 Sequence rules, 94–95 Synonym, 178 Sentiment analysis, 200–202 Session, 187, 189 T Sessionization, 189 Target Sigmoid transformation, 23 Sign operator, 60 definition, 35–38 Similarity measure, 177 variable, 87 Site search, 192 Test sample, 71 Test group, 170 quality, 192 Tie strength prediction, 203 report, 192 Timeliness, 152 usage, 192 Time-varying covariates, 106, 116 Six sigma, 204 Tool vendors, 7 Small data sets, 72 Top decile lift, 76 Social filtering, 176 Top-N recommendation, 183 Social media analytics, 3, 195–204 Total data quality management Social network, 215 learning, 123–124, 165 program, 152 metrics, 121–123 Total quality management (TQM), 204 Sociogram, 120 Traffic light indicator approach, 135, Software, 153–155 commercial, 153 137 open-source, 153 Training sample, 45, 71 Sparseness property, 62 Training set, 51 Spaghetti model, 216 Transaction identifier, 87 Sparse data, 177 Transactional data, 14 Spearman’s rank correlation, 147 Transform Specificity, 74 Spider construction, 167 logarithmic, 112 Splitting decision, 42 Trend analysis, 191 Splitting up data set, 71–74 Triangle, 168 SPSS, 153 Truncation, 23 Squashing, 49 t-test, 143–144 Standardizing data, 24 Two-stage model, 52, 55 Statistical performance, 9, 133 Types of data sources, 13–15 Stemming, 201 Stopping criterion, 45 U Stopping decision, 42, 47 U-matrix, 101 Unary rating, 177 Undersampling, 166

232 ▸ INDEX Univariate Return, 190 correlations, 29 Unique, 190 outliers, 20 Visual data exploration, 17–19 Universal approximation, 64 W Universal approximators, 49 W3C, 185 Unstructured data, 14 Weak classifier, 66 Unsupervised learning, 87, 100, 166 Web analytics, 4, 94, 185–195 US Government Accountability Office, Web beacon, 188 Web data collection, 185–188 156 Web KPI, 188–191 Use limitation principle, 156 Web server log analysis, 185 User agent, 186 Weight regularization, 51 User-based collaborative filtering, 176 Weighted average cost of capital, User-item matrix, 177 37 V Weights of evidence, 28–29 Validation sample, 45 Weka, 153 Validation set, 51 White box model, 48 Validation Wilcoxon test, 110 Winner take all learning, 70 out-of-sample, 134 Winsorizing, 23 out-of-sample, out-of-time, 134 Withdrawal inference, 16 out-of-universe, 134 Workflow net, 213 Value-added, 151 Vantage score, 146 Y Variable interactions, 32 Yahoo Search Marketing, 193 Variable selection, 29–32 Vertex, 119 Z Virtual advisor, 184 z-score standardization, 24 Visit, 188 z-scores, 22 Visitors, 190 New, 190