Interpretable deep learning network significantly improves tropical cyclone intensity forecast accuracy

Institute of Atmospheric Physics, Chinese Academy of Sciences

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

The TCI–KAN architecture: (a) Predictor Pruning Optimization Module; (b) Neural Network Optimization Module; (c) Prediction Module.

view more 

Credit: Keyun Li

Predicting tropical cyclones (TCs) accurately is crucial for disaster mitigation and public safety. Although the forecasting accuracy of TC tracks has improved substantially in recent decades, progress in the forecasting of TC intensity remains limited. In recent years, deep  learning methods have shown great potential in TC intensity prediction; however, they still face challenges, including limited interpretability, cumbersome feature engineering, and unreliable real-time operational forecasts.

To overcome these limitations, the research team led by Professor Wei Zhong from the College of Advanced Interdisciplinary Studies, National University of Defense Technology, China, proposes a novel TC intensity prediction framework that integrates Kolmogorov–Arnold networks (KANs) with a dynamic predictor pruning optimization module—namely, TCI–KAN. The overall architecture of TCI–KAN is illustrated in Fig. 1. It consists of three modules: the predictor pruning optimization module, the neural network optimization module, and the prediction module. The research results were recently published in Atmospheric and Oceanic Science Letters.

Testing results demonstrate that the predictor pruning optimization module can effectively select 15 high-impact predictors from 317 predictors. TCI–KAN achieves superior accuracy in 6-h intensity forecasts, with a mean absolute error (MAE) of 2.85 kt. TCI–KAN significantly outperforms the referenced best records by 31%, 13%, and 6% in MAE compared to the official operational forecast, single deep-learning models, and hybrid deep learning models, respectively. In addition, TCI–KAN is suitable for different basins and TC categories. It exhibits higher accuracy and lower uncertainty in the eastern Pacific region compared to other regions, but the prediction error and uncertainty escalate with the increasing intensity of TCs.

"This work extends the application of interpretable deep learning networks to TC intensity prediction and significantly improves the forecasting accuracy. It not only provides a novel technical pathway for TC intensity prediction, but also promotes the development of the forecasting paradigm integrating data-driven and physical mechanism based methods," states Professor Wei Zhong, the corresponding author of the paper.

The first author of the paper is Keyun Li, a master’s student at the National University of Defense Technology. This study was supported by the National Natural Science Foundation of China [grant numbers 42075011 and 42192552].

---

Journal

Atmospheric and Oceanic Science Letters

DOI

10.1016/j.aosl.2025.100694 

Article Title

Tropical cyclone intensity prediction based on Kolmogorov–Arnold networks with predictor pruning optimization

 

Streamlining the consciousness debate, from trees to hermit crabs

Michigan State University

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

Philosophers explore what it means to acknowledge an animal -- or even a tree - has consciousness. 

view more 

Credit: Sue Nichols, Michigan State University

Beyond spirited dinner party debate, establishing which creatures have consciousness matters in terms of animal welfare and conservation policy. A Michigan State University philosophy scholar has added clarity to a messy philosophical debate.

In this month’s journal Biology & Philosophy, PhD candidate Jonah Branding contributes a decision tree that can be applied to questions such as, do fish feel pain when they’re on a hook? Does an ant feel alarm when protecting its colony? Do banana slugs feel anything when they eat dead leaves on the forest floor? Or are these simpler organisms more like stimulus-response machines, which don’t have any mental experience?

“There has been a lot of work on the question of animal consciousness in recent years and claims about consciousness are starting to be taken seriously for more and more organisms,” Branding said. “In the 1990s, there was serious debate over whether chimpanzees are conscious. Today, there is serious debate over whether plants are conscious.”

Acknowledging that creatures have feelings, thoughts and/or a first-person perspective is messy. In the paper “Can a marker approach exclude?” Branding sets some order to different approaches to ethical, scientific, and philosophical questions.

"Consciousness science is notorious for its twists and turns,” said Kristin Andrews, professor of philosophy at City University of New York and York Research Chair in Animal Minds. “In his paper, Branding offers a road map to help us answer some of the most difficult questions about which beings are conscious."

Andrews who is not involved in the study, was an original signer of the 2024 New York Declaration on Animal Consciousness.

Branding works with markers - observable features of organisms that are thought to correlate with consciousness. They can include specific brain regions, patterns of complex behavior, or sophisticated cognitive capacities. If an animal has a lot of the right kinds of markers, the animal is probably conscious.

But what if an animal shows few or no markers? Branding breaks beliefs about that into two groups. On the side is what he calls symmetry, which is those who need evidence – a right number of markers.  If the creature falls short, it’s not a conscious being.

Those who believe markers are evidence of consciousness but say that an absence of markers doesn’t prove against consciousness are in the asymmetry camp.

Branding, a member of MSU’s Ecology, Evolution, and Behavior Program, nails this as the sticky widget of the disagreement and tries to untangle it by asking what it is about that missing marker that is so important in the first place. He creates a decision tree to help narrow down the questions.

He invites considering the hermit crab. Crabs fail in a lot of markers – lacking sizable brains and generally lousy at showing those sophisticated cognitive capacities. But those crabs select their own shell houses, and when confronted with a maze, they’ve been known to swap out shells to better navigate.

“I argue that, depending on which of these approaches you start with, and what you think about a few other related issues in the field, we can figure out where you should fall on the question of exclusion,” Branding said, adding that the questions help parse out how beings in the natural world should be valued and treated.

 “These are questions of how we interact with the world,” he said. “There are major ethical implications about who we have to care about morally.”

---

Journal

Biology & Philosophy

DOI

10.1007/s10539-025-09989-x 

Article Title

Can a marker approach exclude?

Article Publication Date

8-Oct-2025

 

Emergency Medicaid spending for undocumented immigrants in the US

JAMA Network

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

About The Study: 

In this nationwide analysis, emergency Medicaid accounted for only a small fraction of total Medicaid expenditures. Although states with larger undocumented populations spent approximately 15 times more per capita, emergency Medicaid still constituted less than 1% of overall Medicaid spending even in states with large undocumented populations, posing a limited fiscal burden on Medicaid. These results suggest that cuts to emergency Medicaid will produce minimal overall cost savings and will disproportionately harm states with large undocumented populations.

Corresponding Author: To contact the corresponding author, Patricia Mae G. Santos, MD, MS, email patricia.mae.garcia.santos@emory.edu.

To access the embargoed study: Visit our For The Media website at this link https://media.jamanetwork.com/

(doi:10.1001/jama.2025.18709)

Editor’s Note: Please see the article for additional information, including other authors, author contributions and affiliations, conflict of interest and financial disclosures, and funding and support.

#  #  #

Embed this link to provide your readers free access to the full-text article 

https://jamanetwork.com/journals/jama/fullarticle/10.1001/jama.2025.18709?guestAccessKey=1a4a83ae-59d5-400f-8b9f-4797ee43e18a&utm_source=for_the_media&utm_medium=referral&utm_campaign=ftm_links&utm_content=tfl&utm_term=100925

---

Journal

JAMA

 

BfR Consumer Monitor: Not many people are concerned about raw milk

BfR Federal Institute for Risk Assessment

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

What it's about:

Raw milk (untreated milk) enjoys a certain popularity, but poses health risks. To avoid food poisoning, it should be boiled before drinking. But how well is raw milk actually known among the population? How do people perceive its health risk? Two-thirds of those surveyed had already heard of raw milk. Only ten percent expressed concern, making raw milk the topic of least concern of the risks surveyed. This is according to the BfR Consumer Monitor, a regularly conducted representative survey by the German Federal Institute for Risk Assessment (BfR). “The health risks of raw milk are underestimated,” explains BfR President Professor Andreas Hensel. “It can transmit pathogens. Particularly sensitive population groups such as children, pregnant women, the elderly and sick people should generally avoid unboiled raw milk.” However, even healthy adults face an increased risk of mild to severe food poisoning when consuming unboiled raw milk, depending on the pathogen. Like approved food dyes and mercury in fish, raw milk was a topic covered by the BfR Consumer Monitor for the first time. 

BfR Consumer Monitor 05/2025

“In your opinion, what are the greatest health risks in everyday life, for example in relation to food, personal care products, clothing or toys?” Public opinion is consistent on this open-ended question. The most frequently mentioned are usually undesirable substances (such as “chemicals” or “pollutants”, accounting for 41 percent of responses). These are followed by plastics (17 percent), food additives (10 percent), certain nutrients such as fat, sugar or salt (10 percent), and unhealthy diet and lifestyles (10 percent).

“How concerned are you personally about the following consumer health issues?” Among the predefined response options, the topic of ”microplastics in food” leads the way (67 percent very concerned, 15 percent moderately concerned), followed by “antibiotic resistance” (63 percent and 15 percent, respectively) and “pesticide residues in food” (50 percent and 20 percent, respectively). Here, too, the views of those surveyed remain largely unchanged. 

In addition, for the first time, the current survey asked about overconsumption rather than insufficient intake of vitamins and minerals. The latest data shows that 26 percent feel well informed about excessive intake of vitamins and minerals, while in the previous survey 37 percent of respondents felt well informed about insufficient intake.

Questions and answers on the subject of raw milk on the BfR website:

https://www.bfr.bund.de/en/service/frequently-asked-questions/topic/avoiding-infections-what-should-be-considered-when-consuming-raw-milk/

About the BfR

The German Federal Institute for Risk Assessment (BfR) is a scientifically independent public health institution within the portfolio of the German Federal Ministry of Agriculture, Food and Regional Identity (BMLEH). The BfR advises the Federal Government and the States (‘Laender’) on questions of food, feed, chemical and product safety. The BfR conducts independent research on topics that are closely linked to its assessment tasks.