Which method predicts recidivism best?: a comparison of statistical, machine learning and data mining predictive models

By N. Tollenaar, P. V. D. van der Heijden. 2013

143 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Genetics, Sociology and Political Science, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Cynthia Rudin
2. Berk Ustun
3. Lesia Semenova
4. Jiaming Zeng
5. G. Duwe
6. KiDeuk Kim
7. Tong Wang
8. Chudi Zhong

In Genetics

1. Lesia Semenova
2. G. Duwe
3. Jiaming Zeng
4. Berk Ustun
5. Cynthia Rudin
6. KiDeuk Kim

In Sociology and Political Science

1. Lesia Semenova
2. G. Duwe
3. Jiaming Zeng
4. Berk Ustun
5. Cynthia Rudin
6. KiDeuk Kim
7. Adam Lankford

In Evaluation & Benchmarks

1. Lesia Semenova
2. Berk Ustun
3. G. Duwe
4. Jiaming Zeng
5. Cynthia Rudin
6. KiDeuk Kim

Citations in the corpus

• Please Stop Explaining Black Box Models for High Stakes Decisions (2018, neutral): “…machine learning tools seem to be approximately equally accurate for predicting recidivism, even if we define recidivism in many different ways [Zeng et al., 2017, Tollenaar and van der Heijden, 2013], based on predictions of different types of crimes using many different machine learning methods.”
• Interpretable classification models for recidivism prediction (2015, methodology): “This discussion is captured nicely by Bushway (2013), who described a discrepancy between papers of Berk and Bleich (2013) and Tollenaar and van der Heijden (2013).”
• Sacrificing Accuracy for Transparency in Recidivism Risk Assessment: The Impact of Classification Method on Predictive Performance (2016, methodology): “Of the seven prior studies, three found little di ff erence between logistic regression and machine learning algorithms in predicting recidivism (Hamilton, Neuilly, Lee, & Barnoski, 2015; Liu, Yang, Ramsey, Li, & Cold, 2011; Tollenaar & van der Heijden, 2013).”
• A study in Rashomon curves and volumes: A new perspective on generalization and model simplicity in machine learning (2019, result): “The above vision examples with relatively high accuracy are in contrast with criminal recidivism prediction, where many different machine learning algorithms have essentially identical performance on many different recidivism prediction problems [48, 41, 1], where performance is measured across the ”