An analysis of 500 S&P 500 companies revealed that traditional sector labels only partially explain firms’ financial structures. Machine-learning models predicted sectors from accounting ratios, achieving a maximum validation accuracy of 49.3%. However, an unsupervised analysis identified nine financially interpretable groups that crossed sector boundaries. This suggests that, when comparing profitability, leverage, liquidity, and cash-generation patterns, accounting-based peer groups can complement sectors.

Scientists have developed a new understanding of how fish interact as they school in groups. Fish were found to copy their neighbors to align their movements within schools. The discovery opens a door to examining how the brains of individuals interact to produce group-level collective behavior.

The brain may “tag” experiences during learning to better consolidate the memory of them during sleep, according to a study published August 27^th in the open access journal PLOS Biology by Dan Denis from the University of York, United Kingdom, and colleagues.

A machine-learning tool built from Swedish national health registry data can predict hip fracture risk with high accuracy and no in-person assessment, and identifies far more at-risk individuals than current clinical screening practices, according to a study published August 27^th in the open access journal PLOS Medicine by Kristian Axelsson and Mattias Lorentzon of the University of Gothenburg, Sweden, and colleagues.