Multivariate decision trees are powerful machine learning tools for classification and regression that attract many researchers and industry professionals. An optimal binary tree has two types of vertices, (i) branching vertices which have exactly two children and where datapoints are assessed on a set of discrete features and (ii) leaf vertices at which datapoints are given a prediction, and can be obtained by solving a biobjective optimization problem that seeks to (i) maximize the number of correctly classified datapoints and (ii) minimize the number of branching vertices. Branching vertices are linear combinations of training features and therefore can be thought of as hyperplanes. In this paper, we propose two cut-based mixed integer linear optimization (MILO) formulations for designing optimal binary classification trees (leaf vertices assign discrete classes). Our models leverage on-the-fly identification of minimal infeasible subsystems (MISs) from which we derive cutting planes that hold the form of packing constraints. We show theoretical improvements on the strongest flow-based MILO formulation currently in the literature and conduct experiments on publicly available datasets to show our models' ability to scale, strength against traditional branch and bound approaches, and robustness in out-of-sample test performance. Our code and data are available on GitHub.

Decision trees are powerful tools for classification and regression that attract many researchers working in the burgeoning area of machine learning. One advantage of decision trees over other methods is their interpretability, which is often preferred over other higher accuracy methods that are relatively uninterpretable. A binary classification tree has two types of vertices: (i) branching vertices which have exactly two children and where datapoints are assessed on a set of discrete features; and (ii) leaf vertices at which datapoints are given a discrete prediction. An optimal binary classification tree can be obtained by solving a biobjective optimization problem that seeks to (i) maximize the number of correctly classified datapoints and (ii) minimize the number of branching vertices. In this paper, we propose four mixed integer linear optimization (MILO) formulations for designing optimal binary classification trees: two flow-based formulations and two-cut based formulations. We provide theoretical comparisons between our proposed formulations and the strongest flow-based MILO formulation of Aghaei et al. (2021). We conduct experiments on 13 publicly available datasets to show the models' ability to scale and the strength of a biobjective approach using Pareto frontiers. Our code and data are available on GitHub.

Around this time two years ago, David Alston was running around the frigid streets of Chicago attending sneaker release pop-up parties during NBA All-Star Game weekend, trying to get some exposure for his upstart app, Kickstroid. His pre-pandemic objective: letting sneakerheads know how and where to get the latest reissues of various Air Jordans, the newest "Space Jam: A New Legacy"-inspired Nikes from LeBron James, and a relaunched Adidas D Rose 1 from Chicago hoop legend Derrick Rose. The positive feedback Alston got from users in person and on the app left him and his Kickstroid co-founder and college classmate, Nicco Adams, inspired. This weekend on their app, which launched in January 2020, they're tracking to see how the Nike "LeBron 9 Big Bang 2022" reissue will fare as this year's NBA All-Star Game is being held in Cleveland, James' former stomping ground. They say the shoe, with a $210 retail price, has a current "hype rating" of 7.2 (out of 10), a resell value of 4.6, which they also currently estimated at $284 and could price higher.

A Dutch court has demanded that an algorithm-based system used by the government to identify and track down potential housing and benefit cheats is dropped with immediate effect. As reported by DutchNews, on Wednesday, the District Court of The Hague ruled that the system conflicts with EU human rights and privacy protections. Dubbed System Risk Indication (SyRI), the automatic, machine-learning (ML) tool was used by local Dutch authorities to draw up profiles and lists of individuals suspected of being at high risk of conducting benefits fraud. According to the publication, SyRI creates risk profiles from individuals that committed social security fraud in the past and then scans for "similar" citizen profiles, creating leads for potential investigations into others that may also be committing fraud, or be of a high risk of doing so in the future. SyRI's pooling of citizen data, otherwise kept in separate silos, gave authorities wide-ranging powers and "has been exclusively targeted at neighborhoods with mostly low-income and minority residents," according to UN human rights and poverty rapporteur Philip Alston.

If it seems like every week there's a new scandal about ethics and the tech industry, it's not your imagination. Even as the tech industry is trying to establish concrete practices and institutions around tech ethics, hard lessons are being learned about the wide gap between the practices of "doing ethics" and what people think of as "ethical". This helps explain, in part, why it raises eyebrows when Google dissolves its short-lived AI ethics advisory board, in the face of public outcry about including a controversial alumnus of the Heritage Foundation on it, or when organized pressure from Google's engineering staff results in the cancellation of military contracts. This gap is important, because alongside these decidedly bad calls by those leading the charge for ethics in industry, we are also seeing the tech sector begin investing meaningful resources in the organizational capacity to identify, track, and mitigate the consequences of algorithmic technologies. We are at a point where it would seem that the academics and critics who had exhorted the industry to make such considerations for decades should be declaring a small victory.

Nations around the world are "stumbling zombie-like into a digital welfare dystopia" in which artificial intelligence and other technologies are used to target, surveil and punish the poorest people, the United Nation's monitor on poverty has warned. Philip Alston, UN rapporteur on extreme poverty, has produced a devastating account of how new digital technologies are revolutionizing the interaction between governments and the most vulnerable in society. In what he calls the rise of the "digital welfare state", billions of dollars of public money is now being invested in automated systems that are radically changing the nature of social protection. Alston's report on the human rights implications of the shift will be presented to the UN general assembly on Friday. It says that AI has the potential to improve dramatically the lives of disadvantaged communities, but warns that such hope is being lost amid the constant drive for cost cutting and "efficiency".

Over the past six years, the New York City police department has compiled a massive database containing the names and personal details of at least 17,500 individuals it believes to be involved in criminal gangs. The effort has already been criticized by civil rights activists who say it is inaccurate and racially discriminatory. "Now imagine marrying facial recognition technology to the development of a database that theoretically presumes you're in a gang," Sherrilyn Ifill, president and director-counsel of the NAACP Legal Defense fund, said at the AI Now Symposium in New York last Tuesday. Lawyers, activists, and researchers emphasize the need for ethics and accountability in the design and implementation of AI systems. But this often ignores a a couple of tricky questions: who gets to define those ethics, and who should enforce them?

Artificial intelligence could help a brokerage figure out how likely a prospective client is to actually buy insurance. At least that's what officials with Surex Direct are hoping. Magreth, Alta.-based Surex Direct places home, auto and commercial lines and says it can deliver 10 quotes online in 10 minutes or less. If a person gives Surex information to generate a home or auto quote, the brokerage defines that person as a lead, said Matt Alston, co-founder and chief operating officer of Surex. "We are trying to build out sophistication in scoring leads," Alston said in an interview.