Last Seen Blogs
olivergreengrass
Oliver’s Anecdotes
foodlesoodlesdoodles
OodlesofDoodles
hellscythearts
Art ADHD
ilostmyhouse
im scared
Link
0 notes
Link
esearchers have developed a technique that could help this spacecraft land safely. Their approach can enable an autonomous vehicle to plot a provably safe trajectory in highly uncertain situations where there are multiple uncertainties regarding environmental conditions and objects the vehicle could collide with.
0 notes
Link
Energy storage important to creating affordable, reliable, deeply decarbonized electricity systems
0 notes
Text
UF veterinary students gain shelter medicine skills through clerkship
Thanks to a collaboration between the University of Florida College of Veterinary Medicine and Alachua County, students at the college have gained an opportunity for hands-on learning in the field of shelter medicine and the college has enhanced the offerings of its shelter program to better serve the local community.
College administrators approached the county in the spring of 2020 about the possibility of creating a student clerkship at the Animal Resources & Care, or ARC, shelter after the rotation that had been in operation for five years at Miami-Dade Animal Services was paused due to COVID-19.
Simone Guerios, D.V.M., a clinical assistant professor of shelter medicine who had led the Miami-Dade clerkship, was also interested in relocating to the Alachua County area, said Christopher Adin, D.V.M., a professor of veterinary surgery and chair of the college’s department of small animal clinical sciences, so the UF team reached out to county shelter administrators to ask if they would be willing to house the rotation there.
The county was receptive, and the new rotation began in June 2020.
“Very shortly after the rotation was resumed, all students who had previously been scheduled for (Miami-Dade) instead reported to the Alachua County shelter, where they were able to get an excellent education experience,” Adin said.
In 2021, 116 students trained at the Alachua County shelter.
“Under my supervision, these students performed 1,100 spays and neuters, including 635 dogs and 475 cats, and performed an average of three medical exams per day,” Guerios said. “In addition, they participated in many special surgeries to treat animals with a variety of medical conditions, including eye surgeries, mass removals, hernia repairs, mastectomies and more.”
Students also handled medical examinations, skin scrapes, heartworm tests and treatments, cytology, neonatal care, ophthalmology exams and radiographs, expanding the care that can be provided to ill and injured animals at the shelter.
Guerios said working with the veterinary students and shelter staff at ARC was rewarding, and that four students who previously took the rotation as third-year students returned as seniors to take it again. “They did a great job the first time they were on this rotation, but I was impressed with how amazing they were in the second round,” Guerios said.
Ed Williams, ARC’s director, said the shelter staff loved having the students’ help, and that the county valued its continuing partnership with UF.
“The whole shelter program at the university and the clerkship specifically have allowed us to enhance the level of care we are able to provide to our shelter pets, while helping to train the next generation of veterinarians,” he said.
Veterinary students are thrilled with the opportunity to serve the local community and its unowned animals in Alachua County, Adin added.
“We are also pleased that our Miami-Dade rotation resumed in January,” he said. “Between those two programs and our Veterinary Community Outreach Program, which provide veterinary care to animals from shelters in outlying counties, our college employs five full-time veterinarians who work solely to support the care of unowned animals while they are training veterinary students.”
Through these three programs, the college provides a huge community service, Adin said.
“With three opportunities students can pick from, we offer one of the most robust combined training opportunities in the country to veterinary students interested in shelter medicine,” Adin said. “All of these programs represent a ‘win-win’ where students get hands-on experience and animals receive high-quality care under the guidance of experts in their field.”
Media contact: Sarah Carey at [email protected] or 352-294-4242
https://www.sourcearu.com/UFH/article/uf-veterinary-students-gain-shelter-medicine-skills-through-clerkship/
1 note
·
View note
Text
Unpacking black-box models
Modern machine-learning models, such as neural networks, are often referred to as “black boxes” because they are so complex that even the researchers who design them can’t fully understand how they make predictions.
To provide some insights, researchers use explanation methods that seek to describe individual model decisions. For example, they may highlight words in a movie review that influenced the model’s decision that the review was positive.
But these explanation methods don’t do any good if humans can’t easily understand them, or even misunderstand them. So, MIT researchers created a mathematical framework to formally quantify and evaluate the understandability of explanations for machine-learning models. This can help pinpoint insights about model behavior that might be missed if the researcher is only evaluating a handful of individual explanations to try to understand the entire model.
“With this framework, we can have a very clear picture of not only what we know about the model from these local explanations, but more importantly what we don’t know about it,” says Yilun Zhou, an electrical engineering and computer science graduate student in the Computer Science and Artificial Intelligence Laboratory (CSAIL) and lead author of a paper presenting this framework.
Zhou’s co-authors include Marco Tulio Ribeiro, a senior researcher at Microsoft Research, and senior author Julie Shah, a professor of aeronautics and astronautics and the director of the Interactive Robotics Group in CSAIL. The research will be presented at the Conference of the North American Chapter of the Association for Computational Linguistics.
Understanding local explanations
One way to understand a machine-learning model is to find another model that mimics its predictions but uses transparent reasoning patterns. However, recent neural network models are so complex that this technique usually fails. Instead, researchers resort to using local explanations that focus on individual inputs. Often, these explanations highlight words in the text to signify their importance to one prediction made by the model.
Implicitly, people then generalize these local explanations to overall model behavior. Someone may see that a local explanation method highlighted positive words (like “memorable,” “flawless,” or “charming”) as being the most influential when the model decided a movie review had a positive sentiment. They are then likely to assume that all positive words make positive contributions to a model’s predictions, but that might not always be the case, Zhou says.
The researchers developed a framework, known as ExSum (short for explanation summary), that formalizes those types of claims into rules that can be tested using quantifiable metrics. ExSum evaluates a rule on an entire dataset, rather than just the single instance for which it is constructed.
Using a graphical user interface, an individual writes rules that can then be tweaked, tuned, and evaluated. For example, when studying a model that learns to classify movie reviews as positive or negative, one might write a rule that says “negation words have negative saliency,” which means that words like “not,” “no,” and “nothing” contribute negatively to the sentiment of movie reviews.
Using ExSum, the user can see if that rule holds up using three specific metrics: coverage, validity, and sharpness. Coverage measures how broadly applicable the rule is across the entire dataset. Validity highlights the percentage of individual examples that agree with the rule. Sharpness describes how precise the rule is; a highly valid rule could be so generic that it isn’t useful for understanding the model.
Testing assumptions
If a researcher seeks a deeper understanding of how her model is behaving, she can use ExSum to test specific assumptions, Zhou says.
If she suspects her model is discriminative in terms of gender, she could create rules to say that male pronouns have a positive contribution and female pronouns have a negative contribution. If these rules have high validity, it means they are true overall and the model is likely biased.
ExSum can also reveal unexpected information about a model’s behavior. For example, when evaluating the movie review classifier, the researchers were surprised to find that negative words tend to have more pointed and sharper contributions to the model’s decisions than positive words. This could be due to review writers trying to be polite and less blunt when criticizing a film, Zhou explains.
“To really confirm your understanding, you need to evaluate these claims much more rigorously on a lot of instances. This kind of understanding at this fine-grained level, to the best of our knowledge, has never been uncovered in previous works,” he says.
“Going from local explanations to global understanding was a big gap in the literature. ExSum is a good first step at filling that gap,” adds Ribeiro.
Extending the framework
In the future, Zhou hopes to build upon this work by extending the notion of understandability to other criteria and explanation forms, like counterfactual explanations (which indicate how to modify an input to change the model prediction). For now, they focused on feature attribution methods, which describe the individual features a model used to make a decision (like the words in a movie review).
In addition, he wants to further enhance the framework and user interface so people can create rules faster. Writing rules can require hours of human involvement — and some level of human involvement is crucial because humans must ultimately be able to grasp the explanations — but AI assistance could streamline the process.
As he ponders the future of ExSum, Zhou hopes their work highlights a need to shift the way researchers think about machine-learning model explanations.
“Before this work, if you have a correct local explanation, you are done. You have achieved the holy grail of explaining your model. We are proposing this additional dimension of making sure these explanations are understandable. Understandability needs to be another metric for evaluating our explanations,” says Zhou.
This research is supported, in part, by the National Science Foundation.
https://www.sourcearu.com/MIT/article/machine-learning-explainability-0505/
0 notes
Text
Search reveals eight new sources of black hole echoes
new study, the team developed search algorithm to comb through data taken by NASA’s Neutron star Interior Composition Explorer, or NICER, a high-time-resolution X-ray telescope aboard the International Space Station. The algorithm picked out 26 black hole X-ray binary systems that were previously known to emit X-ray outbursts. Of these 26, the team found that 10 systems were close and bright enough that they could discern X-ray echoes amid the outbursts. Eight of the 10 were previously not known to emit echoes.
“We see new signatures of reverberation in eight sources,” Wang says. “The black holes range in mass from five to 15 times the mass of the sun, and they’re all in binary systems with normal, low-mass, sun-like stars.”
As a side project, Kara is working with MIT education and music scholars, Kyle Keane and Ian Condry, to convert the emission from a typical X-ray echo into audible sound waves. Take a listen to the sound of a black hole echo here:
The researchers then ran the algorithm on the 10 black hole binaries and divided the data into groups with similar “spectral timing features,” that is, similar delays between high-energy X-rays and reprocessed echoes. This helped to quickly track the change in X-ray echoes at every stage during a black hole’s outburst.
The team identified a common evolution across all systems. In the initial “hard” state, in which a corona and jet of high-energy particles dominates the black hole’s energy, they detected time lags that were short and fast, on the order of milliseconds. This hard state lasts for several weeks. Then, a transition occurs over several days, in which the corona and jet sputter and die out, and a soft state takes over, dominated by lower-energy X-rays from the black hole’s accretion disk.
During this hard-to-soft transition state, the team discovered that time lags grew momentarily longer in all 10 systems, implying the distance between the corona and disk also grew larger. One explanation is that the corona may briefly expand outward and upward, in a last high-energy burst before the black hole finishes the bulk of its stellar meal and goes quiet.
“We’re at the beginnings of being able to use these light echoes to reconstruct the environments closest to the black hole,” Kara says. “Now we’ve shown these echoes are commonly observed, and we’re able to probe connections between a black hole’s disk, jet, and corona in a new way.”
This research was supported, in part, by NASA.
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes
Link
0 notes