Cities around the world face a critical shortage of affordable and decent housing. Despite its critical importance for policy, our ability to effectively monitor and track progress in urban housing is limited. Deep learning-based computer vision methods applied to street-level images have been successful in the measurement of socioeconomic and environmental inequalities but did not fully utilize temporal images to track urban change as time-varying labels are often unavailable. We used self-supervised methods to measure change in London using 15 million street images taken between 2008 and 2021. Our novel adaptation of Barlow Twins, Street2Vec, embeds urban structure while being invariant to seasonal and daily changes without manual annotations. It outperformed generic embeddings, successfully identified point-level change in London's housing supply from street-level images, and distinguished between major and minor change. This capability can provide timely information for urban planning and policy decisions toward more liveable, equitable, and sustainable cities.
相關內容
We import the algebro-geometric notion of a complete collineation into the study of maximum likelihood estimation in directed Gaussian graphical models. A complete collineation produces a perturbation of sample data, which we call a stabilisation of the sample. While a maximum likelihood estimate (MLE) may not exist or be unique given sample data, it is always unique given a stabilisation. We relate the MLE given a stabilisation to the MLE given original sample data, when one exists, providing necessary and sufficient conditions for the MLE given a stabilisation to be one given the original sample. For linear regression models, we show that the MLE given any stabilisation is the minimal norm choice among the MLEs given an original sample. We show that the MLE has a well-defined limit as the stabilisation of a sample tends to the original sample, and that the limit is an MLE given the original sample, when one exists. Finally, we study which MLEs given a sample can arise as such limits. We reduce this to a question regarding the non-emptiness of certain algebraic varieties.
The lottery is a very lucrative industry. Popular fascination often focuses on the largest prizes. However, less attention has been paid to detecting unusual lottery buying behaviors at lower stakes. Our paper introduces a new model to detect illegal discounting in the North Carolina Education Lottery using statistical analysis of net gains and ticket buying habits. Nine outlying players are flagged and are further examined using a proposed stochastic model to calculate the range of their possible losses in the lottery. The unusual buying patterns of the players flagged as outliers are further confirmed using a K-means clustering analysis of lottery store visiting behaviors.
The replicator equation in evolutionary game theory describes the change in a population's behaviors over time given suitable incentives. It arises when individuals make decisions using a simple learning process - imitation. A recent emerging framework builds upon this standard model by incorporating game-environment feedback, in which the population's actions affect a shared environment, and in turn, the changing environment shapes incentives for future behaviors. In this paper, we investigate game-environment feedback when individuals instead use a boundedly rational learning rule known as logit learning. We characterize the resulting system's complete set of fixed points and their local stability properties, and how the level of rationality determines overall environmental outcomes in comparison to imitative learning rules. We identify a large parameter space for which logit learning exhibits a wide range of dynamics as the rationality parameter is increased from low to high. Notably, we identify a bifurcation point at which the system exhibits stable limit cycles. When the population is highly rational, the limit cycle collapses and a tragedy of the commons becomes stable.
Direct reciprocity is a mechanism for the evolution of cooperation in repeated social interactions. According to this literature, individuals naturally learn to adopt conditionally cooperative strategies if they have multiple encounters with their partner. Corresponding models have greatly facilitated our understanding of cooperation, yet they often make strong assumptions on how individuals remember and process payoff information. For example, when strategies are updated through social learning, it is commonly assumed that individuals compare their average payoffs. This would require them to compute (or remember) their payoffs against everyone else in the population. To understand how more realistic constraints influence direct reciprocity, we consider the evolution of conditional behaviors when individuals learn based on more recent experiences. Even in the most extreme case that they only take into account their very last interaction, we find that cooperation can still evolve. However, such individuals adopt less generous strategies, and they tend to cooperate less often than in the classical setup with average payoffs. Interestingly, once individuals remember the payoffs of two or three recent interactions, cooperation rates quickly approach the classical limit. These findings contribute to a literature that explores which kind of cognitive capabilities are required for reciprocal cooperation. While our results suggest that some rudimentary form of payoff memory is necessary, it already suffices to remember a few interactions.
Humans have a powerful and mysterious capacity to reason. Working through a set of mental steps enables us to make inferences we would not be capable of making directly even though we get no additional data from the world. Similarly, when large language models generate intermediate steps (a chain of thought) before answering a question, they often produce better answers than they would directly. We investigate why and how chain-of-thought reasoning is useful in language models, testing the hypothesis that reasoning is effective when training data consists of overlapping local clusters of variables that influence each other strongly. These training conditions enable the chaining of accurate local inferences to estimate relationships between variables that were not seen together in training. We prove that there will exist a "reasoning gap", where reasoning through intermediate variables reduces bias, for the simple case of an autoregressive density estimator trained on local samples from a chain-structured probabilistic model. We then test our hypothesis experimentally in more complex models, training an autoregressive language model on samples from Bayes nets but only including a subset of variables in each sample. We test language models' ability to match conditional probabilities with and without intermediate reasoning steps, finding that intermediate steps are only helpful when the training data is locally structured with respect to dependencies between variables. The combination of locally structured observations and reasoning is much more data-efficient than training on all variables. Our results illustrate how the effectiveness of reasoning step by step is rooted in the local statistical structure of the training data.
Existing emotion prediction benchmarks contain coarse emotion labels which do not consider the diversity of emotions that an image and text can elicit in humans due to various reasons. Learning diverse reactions to multimodal content is important as intelligent machines take a central role in generating and delivering content to society. To address this gap, we propose Socratis, a societal reactions benchmark, where each image-caption (IC) pair is annotated with multiple emotions and the reasons for feeling them. Socratis contains 18K free-form reactions for 980 emotions on 2075 image-caption pairs from 5 widely-read news and image-caption (IC) datasets. We benchmark the capability of state-of-the-art multimodal large language models to generate the reasons for feeling an emotion given an IC pair. Based on a preliminary human study, we observe that humans prefer human-written reasons over 2 times more often than machine-generated ones. This shows our task is harder than standard generation tasks because it starkly contrasts recent findings where humans cannot tell apart machine vs human-written news articles, for instance. We further see that current captioning metrics based on large vision-language models also fail to correlate with human preferences. We hope that these findings and our benchmark will inspire further research on training emotionally aware models.
The estimation of the effect of environmental exposures and overall mixtures on a survival time outcome is common in environmental epidemiological studies. While advanced statistical methods are increasingly being used for mixture analyses, their applicability and performance for survival outcomes has yet to be explored. We identified readily available methods for analyzing an environmental mixture's effect on a survival outcome and assessed their performance via simulations replicating various real-life scenarios. Using prespecified criteria, we selected Bayesian Additive Regression Trees (BART), Cox Elastic Net, Cox Proportional Hazards (PH) with and without penalized splines, Gaussian Process Regression (GPR) and Multivariate Adaptive Regression Splines (MARS) to compare the bias and efficiency produced when estimating individual exposure, overall mixture, and interaction effects on a survival outcome. We illustrate the selected methods in a real-world data application. We estimated the effects of arsenic, cadmium, molybdenum, selenium, tungsten, and zinc on incidence of cardiovascular disease in American Indians using data from the Strong Heart Study (SHS). In the simulation study, there was a consistent bias-variance trade off. The more flexible models (BART, GPR and MARS) were found to be most advantageous in the presence of nonproportional hazards, where the Cox models often did not capture the true effects due to their higher bias and lower variance. In the SHS, estimates of the effect of selenium and the overall mixture indicated negative effects, but the magnitudes of the estimated effects varied across methods. In practice, we recommend evaluating if findings are consistent across methods.
In this paper we develop a novel neural network model for predicting implied volatility surface. Prior financial domain knowledge is taken into account. A new activation function that incorporates volatility smile is proposed, which is used for the hidden nodes that process the underlying asset price. In addition, financial conditions, such as the absence of arbitrage, the boundaries and the asymptotic slope, are embedded into the loss function. This is one of the very first studies which discuss a methodological framework that incorporates prior financial domain knowledge into neural network architecture design and model training. The proposed model outperforms the benchmarked models with the option data on the S&P 500 index over 20 years. More importantly, the domain knowledge is satisfied empirically, showing the model is consistent with the existing financial theories and conditions related to implied volatility surface.
This paper does not describe a working system. Instead, it presents a single idea about representation which allows advances made by several different groups to be combined into an imaginary system called GLOM. The advances include transformers, neural fields, contrastive representation learning, distillation and capsules. GLOM answers the question: How can a neural network with a fixed architecture parse an image into a part-whole hierarchy which has a different structure for each image? The idea is simply to use islands of identical vectors to represent the nodes in the parse tree. If GLOM can be made to work, it should significantly improve the interpretability of the representations produced by transformer-like systems when applied to vision or language
Knowledge graphs (KGs) of real-world facts about entities and their relationships are useful resources for a variety of natural language processing tasks. However, because knowledge graphs are typically incomplete, it is useful to perform knowledge graph completion or link prediction, i.e. predict whether a relationship not in the knowledge graph is likely to be true. This paper serves as a comprehensive survey of embedding models of entities and relationships for knowledge graph completion, summarizing up-to-date experimental results on standard benchmark datasets and pointing out potential future research directions.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191