No abstract

No abstract

No abstract

It would be an overt act of omission to permit this decade to expire without reflection on its impact on criminalistics in the United States. Its influences were profound. No prior comparable period matched the growth and maturation of our profession during this decade. In the short span of ten years criminalistics, at first a stepchild of the analytical sciences, grew to achieve full acceptance as a legitimate discipline of applied science. In retrospect, this progress was long overdue considering that criminalistics entered the 1970s not as a new phenomenon but as one that had been in existence for nearly 60 years. Given the political and economic climate of the 1970s, anything less than the unpreceden:ed growth experienced would have warranted a severe condemnation of its professional community.

Preprocessing technique Classification method Autoscale Weighted autoscale Optimum linear transformation I. Linear classifier a) Least-squares 83/83 83/83 83/83 b) Negative feedback 56/83 77/70 74/87 II. 3-Nearest Neighbor 80/60 89/80 92/97 III. Multiclass classifier a) Least-squares 91/87 91/87 91/87 b) Negative feedback 94/90 93/90 78/90method is invariant to all linear transformations, the results are the same for the three preprocessed sets of mass spectral data.The second classification method used in this study was the K-Nearest Neighbor Classification Rule (4) with K equal to three. This method is a multiclass method that does not depend upon linear separability. Hence, classification performance is improved in the last two sets of preprocessed data. The attributes and limitations of this method can be found in the chemical literature (4).The results of the multiclass classifier ( ) introduced in this paper are also found in Table I. Here again, the least squares procedure (a) and the error correction feedback procedure (b) were used to calculate the necessary weight vectors. The multiclass procedure performed very well. The overall performance indicates that the least squares procedure for calculating the weight vector is best. Again, note that least squares solutions are unique and are invariant to all linear transformations of the data. These attributes recommend the least squares multiclass procedure for applications which involve more than two classes. The method is at least as effective as other linear classifiers and comparable in accuracy to the more expensive K-Nearest Neighbor Rule.