This newspaper’s comprehensive foreign media reports recently,Ashesh Rambachan, an expert in the Department of Economics at Harvard University School of Arts and Sciences, published the latest research in the "Quarterly Journal of Economics",Replacing certain judicial decision-making functions with algorithms can eliminate possible systemic biases that bet365 games may have in hearing cases,Thus promoting judicial justice。
Article points out,current,In many cases,Decision makers need to make important choices based on predictions of unknown outcomes。Especially the judge,They decide whether the defendant can be released on bail,Or how to sentence a convicted person。Some companies are already using machine learning-based models in high-stakes decisions。When deploying these learning models,There are various assumptions about human behavior。Other studies have statistically tested these hypothesized behaviors,That is, whether decision makers bet365 games make systematic prediction errors,and developed a method to assess whether their forecasts were systematically biased。
Inspired by the above research, Rambakan began to conduct an in-depth analysis of the pretrial system in New York City, USA。He discovered,Due to defendant’s race、Disclosure of characteristics such as age and criminal history,A considerable number of bet365 games make systematic errors in predicting the risk of defendants’ pretrial misconduct。
Rambakan used public information on bet365 games in New York City in the United States,These bet365 games are randomly assigned to hearing cases in designated courtrooms on a rotating basis,Analysis of the defendant’s release decision without appearing in court。1460462 cases involved,The defendants in 758,027 cases were released early。Lambakan saw,Given the characteristics of the defendant,At least 20% of bet365 games systematically mispredict the risk of defendant misconduct。At least 32% of bet365 games’ decisions are inconsistent with the defendant’s actual ability to post the specified bail。
Lambakan said,Although the impact of replacing judges’ decision-making with algorithmic decision-making rules has not yet been accurately assessed,But it can indeed improve the efficiency of the trial,The probability of systemic bias bet365 games also be greatly reduced。bet365 games algorithms be completely used to replace human decision-makers in the future,Depends on whether humans bet365 games observe the systematic prediction errors they make based on the algorithm,And whether the algorithm is designed with a deep understanding of human behavior。(Zhao Qian/Compiler)
Friendly links:
Website registration number: Beijing Public Network Security No. 11010502030146 Ministry of Industry and Information Technology:
All rights reserved by bet365 games Magazine. No reproduction or use without permission is allowed
Chief editor’s email: zzszbj@126.com Contact information of this website: 010-85886809 Address: Floor 11-12, Building 1, No. 15 Guanghua Road, Chaoyang District, Beijing Postal Code: 100026
>