6/3/2015 T.K. Cocx, Prediction of criminal careers through 2- dimensional Extrapolation W. Kosters et al.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, ?
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, ? ? 2-Dimensional Extrapolation
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Research Area Criminal Career Study Sociology Psychology Criminology Law Computer Science
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Criminal Careers
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Analysis Goal Analysis
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Practical Factors NatureDurationFrequencySeriousness
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Paradigm Four factors Distance Measure Clustering Prediction Strategic analysis done on this
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Alignment
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Calculating Distance between Careers Nature Severity
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Clustering and classification Clustering is done based upon distance Form of multi-dimensional scaling Iterative is necessary After clustering: classes are assigned to visible clusters. By hand 11 classes Classification can be done by k-means
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Results: Clustering and Classification Year 1Year 2Year 3 Year 4
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, 2-Dimensional Extrapolation ? Year 1 Year 2 Year 3 Year 4 The ‘Marble in Funnel’ and the ‘Criminal Career Prediction’ are two variants of the same problem: Extrapolation of a time sequence in a plane.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Regular Mathematical Extrapolation One variable (usually time, x) is given. One Variable (Value, Temperature, weight, etc, y) is dependant on the given variable
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, 2-Dimensional Extrapolation One variable (usually time, t) is given. Two variables (x, y) are dependant on the given variable. Sometimes (as in the criminal career prediction) x and y are meaningless. Only the location relative to already placed elements is important. Relatively under-researched area in mathematics.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Possible solutions 2-Dimensional Extrapolation Assume y depends on x Rotate image to optimally Arrange t-order on x-axis Regular second degree extrapolation Same as Left option Regular third degree extrapolation Assume x depends on t and y depends on t separately Extrapolate separately Combine in {x,y}-system Spline interpolate items t and t+1 Extrapolate after t last Different methods
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Spline Extrapolation There are two choices in spline extrapolation: Straight line cont. Polynomial cont.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Future Class Calculation Select n existing data points closest to extrapolated curve. The closer to ‘last known’ point, the more accurate. Calculate expected attributes of individual under consideration with weighted average of the n points. Classify current individual using these attributes.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Overview of method Two-dimensionalHigh-dimensional Four factors Distance Matrix Crimes committed Clustering Classification Extrapolation Class Prediction Prediction # crimes Combined
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Implementation
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Cluster Reduction
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Results Using the original Dutch National Criminal Record Database (App. 1 million offenders)
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Effects of number of reference points How many reference points are needed? is enough
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Effects of known years How many years should be known for an accurate prediction? 3-5 is enough
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Privacy issues Data mining general truth from lot of data In this case translate this truth to individual cases privacy and statistical issues arise Comparable to data mining on financial transactions Seen as acceptable Reasonably few false positives Operatives familiar with percentages The approach poses no risk to non-offenders only (existing) career continuation
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Conclusion Criminal career analysis can serve as basis for career prediction. Using the concept of 2-dimensional extrapolation on an existing clustering yields the movement in time of an individual from his past to his future Using ‘straight line spline’ extrapolation with the maximum existing elements predicts the future class of an offender with an 88.7% Accuracy.
Prediction of criminal careers through 2-dimensional extrapolation 6/3/2015T.K. Cocx, Interrogation