Goto

Collaborating Authors

 Oceania


NestGNN: A Graph Neural Network Framework Generalizing the Nested Logit Model for Travel Mode Choice

arXiv.org Machine Learning

Nested logit (NL) has been commonly used for discrete choice analysis, including a wide range of applications such as travel mode choice, automobile ownership, or location decisions. However, the classical NL models are restricted by their limited representation capability and handcrafted utility specification. While researchers introduced deep neural networks (DNNs) to tackle such challenges, the existing DNNs cannot explicitly capture inter-alternative correlations in the discrete choice context. To address the challenges, this study proposes a novel concept - alternative graph - to represent the relationships among travel mode alternatives. Using a nested alternative graph, this study further designs a nested-utility graph neural network (NestGNN) as a generalization of the classical NL model in the neural network family. Theoretically, NestGNNs generalize the classical NL models and existing DNNs in terms of model representation, while retaining the crucial two-layer substitution patterns of the NL models: proportional substitution within a nest but non-proportional substitution beyond a nest. Empirically, we find that the NestGNNs significantly outperform the benchmark models, particularly the corresponding NL models by 9.2\%. As shown by elasticity tables and substitution visualization, NestGNNs retain the two-layer substitution patterns as the NL model, and yet presents more flexibility in its model design space. Overall, our study demonstrates the power of NestGNN in prediction, interpretation, and its flexibility of generalizing the classical NL model for analyzing travel mode choice.


Conv4Rec: A 1-by-1 Convolutional AutoEncoder for User Profiling through Joint Analysis of Implicit and Explicit Feedbacks

arXiv.org Artificial Intelligence

We introduce a new convolutional AutoEncoder architecture for user modelling and recommendation tasks with several improvements over the state of the art. Firstly, our model has the flexibility to learn a set of associations and combinations between different interaction types in a way that carries over to each user and item. Secondly, our model is able to learn jointly from both the explicit ratings and the implicit information in the sampling pattern (which we refer to as `implicit feedback'). It can also make separate predictions for the probability of consuming content and the likelihood of granting it a high rating if observed. This not only allows the model to make predictions for both the implicit and explicit feedback, but also increases the informativeness of the predictions: in particular, our model can identify items which users would not have been likely to consume naturally, but would be likely to enjoy if exposed to them. Finally, we provide several generalization bounds for our model, which to the best of our knowledge, are among the first generalization bounds for auto-encoders in a Recommender Systems setting; we also show that optimizing our loss function guarantees the recovery of the exact sampling distribution over interactions up to a small error in total variation. In experiments on several real-life datasets, we achieve state-of-the-art performance on both the implicit and explicit feedback prediction tasks despite relying on a single model for both, and benefiting from additional interpretability in the form of individual predictions for the probabilities of each possible rating.


Apple's Best New iOS 26 Feature Has Been on Pixel Phones for Years

WIRED

Apple's Best New iOS 26 Feature Has Been on Pixel Phones for Years The iPhone's new software screens your calls using machine intelligence. Neat, but Google had the feature first--just like so many other features that rely on AI to work. Call Screening on an iPhone. Ever since I was a child, I've despised answering the phone when an unknown number calls. Who could be on the other end?


What do we know about alleged Trump signature on Epstein letter?

BBC News

What do we know about alleged Trump signature on Epstein letter? A birthday letter to Jeffrey Epstein from 2003, allegedly signed by Donald Trump, has been released by US lawmakers. The White House has said, the president did not write this letter, he did not sign this letter, but Democrats believe it is authentic. There has been a particular focus on the signature - BBC Verify's Jake Horton has been examining it, with the help of handwriting experts. The BBC's Tom Bateman spoke with Patrick Scallen who lives near the Annunciation Church and ran towards the sound of gunfire.


Does Society Have Too Many Rules?

The New Yorker

Does Society Have Too Many Rules? When regular people seem burdened by bureaucracy, and the powerful act as they choose, it's worth asking whether we've forgotten what makes rules effective. I live in a three-generation household. Our place is big, but crowded: all of us have hobbies, and so every shelf or surface contains toys, books, art supplies, sporting goods, craft projects, cameras, musical instruments, or kitchen gadgets. Before the table can be set for dinner, it must be cleared of a board game or marble run. My desk, where I aim to write in the mornings, has been repurposed as a drone-repair workshop. The property includes two broken-down sheds and a garage.


Is AI the New Frontier of Women's Oppression?

WIRED

Is AI the New Frontier of Women's Oppression? In her new book, feminist author Laura Bates explores how sexbots, AI assistants, and deepfakes are reinventing misogyny and harming women. After spending her early twenties as a nanny in the UK, Laura Bates noticed that the young girls she was caring for were preoccupied by their bodies, spurred on by the marketing they were receiving. In 2012, Bates, a London-based feminist author and activist, started The Everyday Sexism Project, a website dedicated to documenting and combatting sexism, misogyny, and gendered violence around the world by highlighting insidious instances of it such as invisible labor, referring to women as girls and commenting on their attire in professional settings. The site was turned into a book in 2014.


How Google dodged a major breakup – and why OpenAI is to thank for it

The Guardian

The reason for the relative tameness of the penalty is the emergence of real competition to Google - what the case concerned in the first place. The reason for the relative tameness of the penalty is the emergence of real competition to Google - what the case concerned in the first place. An antitrust apocalypse has been averted, and it's all down to its biggest competitor, according to the judge who could've forced Google to sell Chrome I'm your host, Blake Montgomery, writing to you as I finish the audiobook version of Don DeLillo's White Noise, which I can't say I found compelling. In tech - artificial intelligence is having its day in court with an 11th-hour appearance in Google's landmark antitrust trial and Anthropic's major settlement with book authors. Google dodged a catastrophic breakup, and it has its biggest competitor to thank for that, according to the judge who could have forced the tech giant to sell off Chrome, the most popular web browser in the world, and perhaps Android, the world's most widely used mobile operating system.


The women in love with AI chatbots: 'I vowed to him that I wouldn't leave him'

The Guardian

'Some people go into AI relationships purposefully, some out of curiosity, and others accidentally.' 'Some people go into AI relationships purposefully, some out of curiosity, and others accidentally.' The women in love with AI chatbots: 'I vowed to him that I wouldn't leave him' The Guardian's journalism is independent. We will earn a commission if you buy something through an affiliate link. A young tattoo artist on a hiking trip in the Rocky Mountains cozies up by the campfire, as her boyfriend Solin describes the constellations twinkling above them: the spidery limbs of Hercules, the blue-white sheen of Vega. The Guardian's journalism is independent. We will earn a commission if you buy something through an affiliate link. Somewhere in New England, a middle-aged woman introduces her therapist to her husband, Ying.


Hasan Piker Will Never Run for Office

WIRED

The Twitch streamer could pivot from influencer to candidate. But he tells WIRED's podcast he'd rather use his platform to tell Dems "you can't podcast your way out of this problem." Hasan Piker is many things to many people. They don't all feel the same way about Piker or his politics, but most presumably agree on one thing: He is a relentless human being. Most days a week, you can find the 34-year-old Twitch streamer talking to his audience, often for six to nine hours at a stretch. And during President Trump's second term, there's plenty of that to go around. He has nearly 3 million followers on Twitch and has hosted conversations with Senator Bernie Sanders and US representative Alexandria Ocasio-Cortez. He claims his election night stream in 2024 reached a staggering 7.5 million viewers. On this episode of, I talked to Piker about his looks, his love of Italian sandwiches, and any future political aspirations he might (or might not) want to tease. It's great to be here. I heard you were just at the gym. Yeah, I was at the park. Some days I take my dog and I play a little bit of basketball and get to hang out with some people.


Inside Kyiv government building hit by missile strike

BBC News

Ukraine's main government building in Kyiv was hit for the first time since Russia's full-scale invasion of the country on Sunday, officials said. The BBC's Sarah Rainsford visited the scene, where she observed a huge amount of damage. Local media reports suggest a cable came loose along the railway's route, causing it to lose control - a national day of mourning is being observed Actor Julia Roberts makes her Venice Film Festival debut promoting her new movie After The Hunt. The helicopter was attempting to collect water to fight wildfires at the time of the crash. 'Give it a go!': Tips from a top rate tree hugger Top tree hugger Hannah Willow explains why she loves the sport so much.