共找到 20 条结果
This Policy Comment describes how the Food Policy article entitled 'Cost and affordability of nutritious diets at retail prices: Evidence from 177 countries' (first published October 2020) and 'Retail consumer price data reveal gaps and opportunities to monitor food systems for nutrition' (first published September 2021) advanced the use of least-cost benchmark diets to monitor and improve food security. Those papers contributed to the worldwide use of least-cost diets as a new diagnostic indicator of food access, helping to distinguish among causes of poor diet quality related to high prices, low incomes, or displacement by other food options, thereby guiding intervention toward universal access to healthy diets.
New methods for modeling least-cost diets that meet nutritional requirements for health have emerged as important tools for informing nutrition policy and programming around the world. This study develops a three-step approach using cost of healthy diet to inform targeted nutrition programming in Indonesia. We combine detailed retail prices and household survey data from Indonesia to describe how reported consumption and expenditure patterns across all levels of household income diverge from least cost healthy diets using items from nearby markets. In this analysis, we examine regional price variations, identify households with insufficient income for healthy diets, and analyze the nutrient adequacy of reported consumption patterns. We find that household food spending was sufficient to meet national dietary guidelines using the least expensive locally available items for over 98% of Indonesians, but almost all households consume substantial quantities of discretionary foods and mixed dishes while consuming too little energy from fruits, vegetables, and legumes, nuts, and seeds. Households with higher incomes have higher nutrient adequacy and are closer to meeting local dietary gui
Using real-world food price and greenhouse gas (GHG) emissions data for locally available food items in 171 countries, we measure how healthy diets could be obtained with the lowest possible emissions, compared to costs and emissions of the least expensive options and foods most commonly consumed. We find that foods with the lowest GHG emissions for a healthy diet would emit 0.67 kg CO2e. A healthy diet using the least expensive items in each country would emit 1.65 kg CO2e and cost 9.96. Ninety-one percent of the difference in emissions between the lowest-cost and lowest-emissions diets is driven by animal-source foods and starchy staples. Other food groups, especially fruits and vegetables, vary widely in cost but not in emissions. Results show how changes in food policy and choice can most cost-effectively support healthier and more sustainable diets worldwide.
Large-scale food fortification (LSFF) is a widely accepted intervention to alleviate micronutrient deficiencies, yet policy implementation is often incomplete and its effects on diet costs are not well established. We estimated the extent to which LSFF reduces the cost of nutrient-adequate diets using retail food prices and fortification policy data from 89 countries. In total, we modeled 5,874 least-cost diets across 22 sex-age groups and 3 nutrient-adequacy scenarios: meeting nutrient requirements only; adding minimum intakes for starchy staples and fruits and vegetables; and aligning food group shares with national consumption patterns. Assuming 90% implementation of existing LSFF standards, we found median cost reductions of 1.7%, 2.4%, and 4.5% across the three scenarios. Cost reductions varied widely by sex-age groups, national fortification strategies and food price structures. These findings highlight that LSFF may improve diet affordability when policies are carefully designed for local contexts, making it a valuable complement to other efforts that improve access to nutritious diets.
Newborns perceive the world with low-acuity, color-degraded, and temporally continuous vision, which gradually sharpens as infants develop. To explore the ecological advantages of such staged "visual diets", we train self-supervised learning (SSL) models on object-centric videos under constraints that simulate infant vision: grayscale-to-color (C), blur-to-sharp (A), and preserved temporal continuity (T)-collectively termed CATDiet. For evaluation, we establish a comprehensive benchmark across ten datasets, covering clean and corrupted image recognition, texture-shape cue conflict tests, silhouette recognition, depth-order classification, and the visual cliff paradigm. All CATDiet variants demonstrate enhanced robustness in object recognition, despite being trained solely on object-centric videos. Remarkably, models also exhibit biologically aligned developmental patterns, including neural plasticity changes mirroring synaptic density in macaque V1 and behaviors resembling infants' visual cliff responses. Building on these insights, CombDiet initializes SSL with CATDiet before standard training while preserving temporal continuity. Trained on object-centric or head-mounted infant v
The cost and affordability of least-cost healthy diets by time and place are increasingly used as a proxy for access to nutrient-adequate diets. Recent work has focused on the nutrient requirements of individuals, although most food and anti-poverty programs target whole households. This raises the question of how the cost of a nutrient-adequate diet can be measured for an entire household. This study identifies upper and lower bounds on the feasibility, cost, and affordability of meeting all household members' nutrient requirements using 2013-2017 survey data from Malawi. Findings show only a minority of households can afford the nutrient-adequate diet at either bound, with 20% of households able to afford the (upper bound) shared diets and 38% the individualized (lower bound) diets. Individualized diets are more frequently feasible with locally available foods (90% vs. 60% of the time) and exhibit more moderate seasonal fluctuation. To meet all members' needs, a shared diet requires a more nutrient-dense combination of foods that is more costly and exhibits more seasonality in diet cost than any one food group or the individualized diets. The findings further help adjudicate the
People with diabetes need insulin delivery to effectively manage their blood glucose levels, especially after meals, because their bodies either do not produce enough insulin or cannot fully utilize it. Accurate insulin delivery starts with estimating the nutrients in meals and is followed by developing a detailed, personalized insulin injection strategy. These tasks are particularly challenging in daily life, especially without professional guidance. Existing solutions usually assume the prior knowledge of nutrients in meals and primarily rely on feedback from professional clinicians or simulators to develop Reinforcement Learning-based models for insulin management, leading to extensive consumption of medical resources and difficulties in adapting the models to new patients due to individual differences. In this paper, we propose DIETS, a novel diabetic insulin management framework built on the transformer architecture, to help people with diabetes effectively manage insulin delivery in everyday life. Specifically, DIETS tailors a Large Language Model (LLM) to estimate the nutrients in meals and employs a titration model to generate recommended insulin injection strategies, which
To support the expected increase in aquaculture production during the next years, a wider range of alternative ingredients to fishmeal is needed, towards contributing to an increase in production sustainability. This study aimed to test diets formulated with non-conventional feed ingredients on gilthead seabream (Sparus aurata) growth performance, feed utilization, apparent digestibility of nutrients and nutrient outputs to the environment. Four isonitrogenous and isoenergetic diets were formulated: a control diet (CTRL) similar to a commercial feed and three experimental diets containing, as main protein sources, plant by-products, glutens and concentrates (PLANT); processed animal proteins (PAP); or micro/macroalgae, insect meals and yeast (EMERG). Diets were tested in triplicate during 80 days. The (EMERG) treatment resulted in lower fish growth performance, higher FCR and lower nutrient and energy retentions than the other treatments. The lowest protein digestibility was found for the EMERG diet, which caused increased nitrogen losses. The PLANT and PAP treatments resulted in better fish growth performance, higher nutrient and energy retentions, and lower FCR than the CTRL trea
The least-cost diet problem introduces students to optimization and linear programming, using the health consequences of food choice. We provide a graphical example, Excel workbook and Word template using actual data on item prices, food composition and nutrient requirements for a brief exercise in which students guess at and then solve for nutrient adequacy at lowest cost, before comparing modeled diets to actual consumption which has varying degrees of nutrient adequacy. The graphical example is a 'three sisters' diet of corn, beans and squash, and the full multidimensional model is compared to current food consumption in Ethiopia. This updated Stigler diet shows how cost minimization relates to utility maximization, and links to ongoing research and policy debates about the affordability of healthy diets worldwide.
Public opinion reflects and shapes societal behavior, but the traditional survey-based tools to measure it are limited. We introduce a novel approach to probe media diet models -- language models adapted to online news, TV broadcast, or radio show content -- that can emulate the opinions of subpopulations that have consumed a set of media. To validate this method, we use as ground truth the opinions expressed in U.S. nationally representative surveys on COVID-19 and consumer confidence. Our studies indicate that this approach is (1) predictive of human judgements found in survey response distributions and robust to phrasing and channels of media exposure, (2) more accurate at modeling people who follow media more closely, and (3) aligned with literature on which types of opinions are affected by media consumption. Probing language models provides a powerful new method for investigating media effects, has practical applications in supplementing polls and forecasting public opinion, and suggests a need for further study of the surprising fidelity with which neural language models can predict human responses.
Diet design for vegetarian health is challenging due to the limited food repertoire of vegetarians. This challenge can be partially overcome by quantitative, data-driven approaches that utilise massive nutritional information collected for many different foods. Based on large-scale data of foods' nutrient compositions, the recent concept of nutritional fitness helps quantify a nutrient balance within each food with regard to satisfying daily nutritional requirements. Nutritional fitness offers prioritisation of recommended foods using the foods' occurrence in nutritionally adequate food combinations. Here, we systematically identify nutritionally recommendable foods for semi- to strict vegetarian diets through the computation of nutritional fitness. Along with commonly recommendable foods across different diets, our analysis reveals favourable foods specific to each diet, such as immature lima beans for a vegan diet as an amino acid and choline source, and mushrooms for ovo-lacto vegetarian and vegan diets as a vitamin D source. Furthermore, we find that selenium and other essential micronutrients can be subject to deficiency in plant-based diets, and suggest nutritionally-desirabl
Background: Higher endogenous testosterone levels are associated with reduced chronic disease risk and mortality. Since the mid-20th century, there have been significant changes in dietary patterns, and men's testosterone levels have declined in western countries. Cross-sectional studies show inconsistent associations between fat intake and testosterone in men. Methods: Studies eligible for inclusion were intervention studies, with minimal confounding variables, comparing the effect of low-fat vs high-fat diets on men's sex hormones. 9 databases were searched from their inception to October 2020, yielding 6 eligible studies, with a total of 206 participants. Random effects meta-analyses were performed using Cochrane's Review Manager software. Cochrane's risk of bias tool was used for quality assessment. Results: There were significant decreases in sex hormones on low-fat vs high-fat diets. Standardised mean differences with 95% confidence intervals (CI) for outcomes were: total testosterone [-0.38 (95% CI -0.75 to -0.01) P = 0.04]; free testosterone [-0.37 (95% CI -0.63 to -0.11) P = 0.005]; urinary testosterone [-0.38 (CI 95% -0.66 to -0.09) P = 0.009], and dihydrotestosterone [-0
With the widespread adoption of social media sites like Twitter and Facebook, there has been a shift in the way information is produced and consumed. Earlier, the only producers of information were traditional news organizations, which broadcast the same carefully-edited information to all consumers over mass media channels. Whereas, now, in online social media, any user can be a producer of information, and every user selects which other users she connects to, thereby choosing the information she consumes. Moreover, the personalized recommendations that most social media sites provide also contribute towards the information consumed by individual users. In this work, we define a concept of information diet -- which is the topical distribution of a given set of information items (e.g., tweets) -- to characterize the information produced and consumed by various types of users in the popular Twitter social media. At a high level, we find that (i) popular users mostly produce very specialized diets focusing on only a few topics; in fact, news organizations (e.g., NYTimes) produce much more focused diets on social media as compared to their mass media diets, (ii) most users' consumptio
Large language models (LLMs) have demonstrated remarkable capabilities, but their massive scale poses significant challenges for practical deployment. Structured pruning offers a promising solution by removing entire dimensions or layers, yet existing methods face critical trade-offs: task-agnostic approaches cannot adapt to task-specific requirements, while task-aware methods require costly training to learn task adaptability. We propose DIET (Dimension-wise global pruning of LLMs via merging Task-wise importance scores), a training-free structured pruning method that combines dimension-level granularity with task-aware selection. DIET profiles activation magnitudes across tasks using only 100 samples per task, then applies majority voting to construct a single global mask. DIET does not require large costs from pre-computation or training. Experiments on seven zero-shot benchmarks using Gemma-2 2B and 9B models demonstrate the effectiveness of DIET; for example, at 20% sparsity on Gemma-2 2B, DIET achieves near 10% average accuracy improvement, compared to previous state-of-the-art structured pruning methods. This advantage persists across various sparsity levels and model scales
Modern deep recommender models are trained under a continual learning paradigm, relying on massive and continuously growing streaming behavioral logs. In large-scale platforms, retraining models on full historical data for architecture comparison or iteration is prohibitively expensive, severely slowing down model development. This challenge calls for data-efficient approaches that can faithfully approximate full-data training behavior without repeatedly processing the entire evolving data stream. We formulate this problem as \emph{streaming dataset distillation for recommender systems} and propose \textbf{DIET}, a unified framework that maintains a compact distilled dataset which evolves alongside streaming data while preserving training-critical signals. Unlike existing dataset distillation methods that construct a static distilled set, DIET models distilled data as an evolving training memory and updates it in a stage-wise manner to remain aligned with long-term training dynamics. DIET enables effective continual distillation through principled initialization from influential samples and selective updates guided by influence-aware memory addressing within a bi-level optimization
Nutrigenomics is an emerging field that explores the intricate interaction between genes and diet. This study aimed to develop a comprehensive database to help clinicians and patients understand the connections between genetic disorders, associated genes, and tailored nutritional recommendations.
Despite years of research and the dramatic scaling of artificial intelligence (AI) systems, a striking misalignment between artificial and human vision persists. Contrary to humans, AI relies heavily on texture-features rather than shape information, lacks robustness to image distortions, remains highly vulnerable to adversarial attacks, and struggles to recognise simple abstract shapes within complex backgrounds. To close this gap, here we take inspiration from how human vision develops from early infancy into adulthood. We quantified visual maturation by synthesising decades of research into a novel developmental visual diet (DVD) for AI vision. Guiding AI systems through this human-inspired curriculum, which considers the development of visual acuity, contrast sensitivity, and colour, produces models that better align with human behaviour on every hallmark of robust vision tested, yielding the strongest reported reliance on shape information to date, abstract shape recognition beyond the state of the art, and higher resilience to image corruptions and adversarial attacks. Our results thus demonstrate that robust AI vision can be achieved by guiding how a model learns, not merely
Multi-objective evolutionary algorithms (MOEAs) are essential for solving complex optimization problems, such as the diet problem, where balancing conflicting objectives, like cost and nutritional content, is crucial. However, most MOEAs focus on optimizing solutions in the objective space, often neglecting the diversity of solutions in the decision space, which is critical for providing decision-makers with a wide range of choices. This paper introduces an approach that directly integrates a Hamming distance-based measure of uniformity into the selection mechanism of a MOEA to enhance decision space diversity. Experiments on a multi-objective formulation of the diet problem demonstrate that our approach significantly improves decision space diversity compared to NSGA-II, while maintaining comparable objective space performance. The proposed method offers a generalizable strategy for integrating decision space awareness into MOEAs.
Reconstructing sharp 3D representations from blurry multi-view images are long-standing problem in computer vision. Recent works attempt to enhance high-quality novel view synthesis from the motion blur by leveraging event-based cameras, benefiting from high dynamic range and microsecond temporal resolution. However, they often reach sub-optimal visual quality in either restoring inaccurate color or losing fine-grained details. In this paper, we present DiET-GS, a diffusion prior and event stream-assisted motion deblurring 3DGS. Our framework effectively leverages both blur-free event streams and diffusion prior in a two-stage training strategy. Specifically, we introduce the novel framework to constraint 3DGS with event double integral, achieving both accurate color and well-defined details. Additionally, we propose a simple technique to leverage diffusion prior to further enhance the edge details. Qualitative and quantitative results on both synthetic and real-world data demonstrate that our DiET-GS is capable of producing significantly better quality of novel views compared to the existing baselines. Our project page is https://diet-gs.github.io
Continued pretraining offers a promising solution for adapting foundation models to a new target domain. However, in specialized domains, available datasets are often very small, limiting the applicability of SSL methods developed for large-scale pretraining and making hyperparameter search infeasible. In addition, pretrained models are usually released as backbone-weights only, lacking important information to continue pretraining. We propose to bridge this gap with DIET-CP, a simple continued pretraining strategy, where any strong foundation model can be steered towards the new data distribution of interest. DIET-CP relies on a very simple objective, requires no labels, and introduces no more hyperparameters than supervised finetuning. It is stable across data modalities and backbone choices, while providing a significant performance boost for state-of-the-art models such as DINOv3 using only 1000 images.