Search

In this paper an Information-Theoretic method for reconstructing noisy and blurry images is developed. Basically, the inverse problem is transformed into a generalized moment problem, which is then solved by an information theoretic method. This estimation approach is robust for a whole class of distributions and allows the use of prior information. The resulting method builds on the foundations of information-theoretic methods, uses minimal distributional assumptions, performs well and uses efficiently all the available information (hard and soft data). This method is computationally efficient. A number of empirical examples are presented. © 2005 American Institute of Physics [ABSTRACT FROM AUTHOR]

This paper will describe in details two specialized analytical procedures that are needed for using observational data to validate a risk-needs screening instrument: data reduction and risk suppression. First, typically, there are numerous items available for inclusion in a risk screening instrument. These may include redundant, irrelevant, or ill-posed information, however. Data reduction is an essential step if we are to make efficient use of all available information. Second, the outcomes used in any analysis will typically not measure latent risk but suppressed risk. As such, specialized analysis has to be conducted to recover from these data the underlying latent risk that one wants to instrument to reflect. The paper will explain these two aspects of the work to validate risk assessment instruments. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

Variations in the intensity of supervision, defined usually as the frequency of community supervision officer contact, is a common tool used by most supervision agencies for monitoring a heterogeneous offender population. Not all offenders are at equal risk of re-offending and supervising them at the same intensity would be an inefficient use of limited resources. Most agencies use risk assessment instruments to classify offenders based on their risk of supervision failure. The Court Services and Offender Supervision Agency (CSOSA) for the District of Columbia uses a tool called the Auto Screener to perform these assessment and classification functions. This paper uses a regression-discontinuity design to assess the marginal contribution of the intensity of supervision, net of underlying risk, in reducing recidivism. The Auto Screener also permits supervising officers to override the prescribed intensity level, thereby permitting us to quantify the public-safety consequences of such overrides. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

The primary goal of this research is to estimate the population of drug-dependent offenders that could be effectively served by partnerships between courts and drug intervention programs. This population estimate would be presented in terms of sex, age, race/ethnicity, and drug of choice. Questions addressed were 1) what proportion of the drug-dependent population could be effectively treated by conventional drug courts? and, 2) if drug court programs were expanded, what would be the size of the treatment effect? Estimates of the potential impact of these programs on crime levels and resulting economic burdens will be presented. he implication and application to the field would be a true estimate as to the size of the population for drug courts if they were allowed to "go to scale" and what the costs and benefits might be for that expansion. There may be a need to re-examine the eligibility requirements for drug court participants. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

If recidivism is defined as re-arrest within a finite period following release from prison, then the three kinds of outcomes typically available to researchers include: (i) whether or not the individual was rearrested within the follow-up period; (ii) how many times the individual was rearrested; and (iii) what was the duration from release to first (or subsequent) re-arrest. Since these outcomes are all different manifestations of the same underlying stochastic process, they provide us multiple analogies from which to recover information about it. This paper develops a semi-parametric approach for utilizing information in these, and several other related outcomes, to predict criminal recidivism. Accuracy is assessed in both out-of-sample and off-the-support prediction problems. Implications of the modeling strategy for actuarial risk assessment work are discussed. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]