To find out which factors might be linked to executions, the researchers first "trained" their [Artificial Neural Network] by entering the profiles of 1000 death row inmates between 1973 and 2000. Half of this sample of prisoners had been executed and the other half had survived. Each profile contained 18 factors, including the inmate's sex, age, race, marital status, educational level and information on their capital offences.The researchers then tried to determine which factors were most salient: gender (virtually no female prisoners are executed) and education--or, more precisely, "the number of years the inmate had spent in high school." It's suggested that prisoners without much education are worse at working on appeals.
They then fed in profiles for 300 more inmates from the same period and asked the ANN to predict what had happened to them. To their astonishment, it correctly predicted the fates of more than 90 per cent of those inmates.
Jun 25, 2008
dead man predicting
A computer can predict executions with 90% accuracy. How? [sub. req.]