Consumption-led economic growth is crucial for enhancing economic resilience, improving social welfare, and fostering endogenous drivers for innovative development. The purpose of this study is to develop a novel multi-criteria evaluation framework to assess the level of consumption-led economic development in the Yangtze River Delta urban agglomeration. Initially, an evaluation system for consumption-led economic development is constructed across five dimensions: economic autonomy, demand structure, consumption level, consumption structure, and consumption environment. The evaluation framework based on the Vertical and Horizontal Scatter Degree and Entropy Method (VHSD-EM), is then applied to analyze the development level of consumption-led patterns in the core cities of the Yangtze River Delta in China from 2015 to 2021. The empirical results reveal significant disparities in consumption-led economic development among the cities. In 2021, Shanghai achieved the highest comprehensive score (7.83), followed by Hangzhou, Suzhou, Hefei, Ningbo, and Nanjing. The average score for the region was 7.37, suggesting that the Yangtze River Delta urban agglomeration is transitioning toward a consumption-led growth model, with some cities exhibiting characteristics of a high-mass consumption stage. However, the development stages vary across cities, reflecting differences in economic structure and policy focus. Finally, several recommendations are suggested based on the numerical analysis.
History of scholarship and learning. The humanities, Social Sciences
Conversation openings are crucial for effective daily communication. However, existing research predominantly focuses on their sequential structures, with little attention given to their linguistic features and variations across different genres. To address this gap, the present study examines conversation openings from four perspectives (i.e., grammatical forms, sentential types, semantic categories, and social actions) by constructing a corpus of English TV dramas. The corpus comprises dialogues from popular TV series shown in recent years with high Internet Movie Database (IMDb) ratings, totaling 530 conversation openings in three generic settings: educational, medical, and judicial. To ensure data accuracy and reliability, the study employed comprehensive data processing and rigorous analysis using software tools such as AntConc 3.5 and IBM SPSS Statistics 26.0. The findings reveal that full sentences are more prevalent than elliptical ones, with declarative and interrogative sentences being the most frequently used. Speakers tend to favor other-oriented utterances, with requests and greetings emerging as the most common social actions. Additionally, the study identifies significant influences of factors such as gender, familiarity, and social status on the structure and use of conversation openings.
History of scholarship and learning. The humanities, Social Sciences
Since 2018, France's centralized higher education platform, Parcoursup, has implemented quotas for scholarship recipients, with program-specific thresholds based on the applicants' composition. Using difference-in-differences methods, I find that these quotas enabled scholarship students to access more selective programs, although the intention-to-treat effects remain modest. Matching methods reveal that the policy improved the scholarship students' waiting list positions relative to those of comparable non-scholarship peers, and simulations suggest that the modest effect could be attributed to the low intensity of the treatment. However, I detect no robust or lasting effects on the extensive margin of higher education access. Despite high policy salience, quotas did not affect the application behavior or pre-college investment of scholarship students, even among high achievers. These findings align with research on affirmative action, suggesting that such policies primarily benefit disadvantaged students who access selective institutions, rather than expanding total enrollment. Nevertheless, scholarship quotas demonstrate that race-neutral alternatives can effectively promote socioeconomic diversity in prestigious programs.
With growing concerns about user data collection, individualized privacy has emerged as a promising solution to balance protection and utility by accounting for diverse user privacy preferences. Instead of enforcing a uniform level of anonymization for all users, this approach allows individuals to choose privacy settings that align with their comfort levels. Building on this idea, we propose an adapted method for enabling Individualized Differential Privacy (IDP) in Federated Learning (FL) by handling clients according to their personal privacy preferences. By extending the SAMPLE algorithm from centralized settings to FL, we calculate client-specific sampling rates based on their heterogeneous privacy budgets and integrate them into a modified IDP-FedAvg algorithm. We test this method under realistic privacy distributions and multiple datasets. The experimental results demonstrate that our approach achieves clear improvements over uniform DP baselines, reducing the trade-off between privacy and utility. Compared to the alternative SCALE method in related work, which assigns differing noise scales to clients, our method performs notably better. However, challenges remain for complex tasks with non-i.i.d. data, primarily stemming from the constraints of the decentralized setting.
This chapter recounts my lived experience as a novice Digital Humanist. It is deliberately anecdotal, rather than theoretical, in style and form. The chapter tells the story of how I commenced a doctorate in the field of book history, then, with minimal technical training, came to build a large relational database that both enabled and complemented my written dissertation as well as providing value for future users. My research is centred on Angus & Robertson, the largest 20th-century Australian bookseller and publishing house. I was particularly interested in Angus & Robertson’s use of book reviews as a promotional tool. The company archive contains millions of miscellaneous documents and, even when limited to certain subsets, there were thousands of undigitised pages to interrogate. In response to that scale, I turned to the Digital Humanities, using the Heurist platform to design a bespoke database schema then populate the requisite fields with metadata from the physical documents, and subsequently enriching the records with secondary research. The resultant Angus & Robertson Book Reviews Database, which has been published online, remains a living database that at the time of writing contains 152,000 records, each with several fields, amounting to over a million data points.In this chapter, I explain design decisions as well as obstacles that I encountered whilst building the database without prior technical skills. I also share how the database has allowed me to tell previously untold stories about Angus & Robertson, book reviewing, and the 20th-century Australian print industry. The chapter concludes with a discussion of the ongoing potential of this specific database and how platforms like Heurist extend important opportunities to novice Digital Humanists.
Perceived cuteness motivates people to approach cute objects, but no evidence exists of unconscious approach behavior toward objects. Given the unconscious responses associated with cuteness perception, an unconscious drive to physically approach cute objects is likely to occur. However, approach behavior may or may not occur depending on whether or not the perceived cute object is an adult, a baby, or a non-human. In this study, we recruited 24 participants and conducted a correlation study between cuteness ratings and body sway to examine whether or not the approach response is dependent on perceived cuteness. Results showed that the approach or avoidance response to cute objects was observed only in women. For babies, the approach response occurred regardless of the degree of cuteness, but for adult female faces, the cuter the face, the more the avoidance response occurred. For non-human images, the cuter the image, the more the approach response occurred only in early period of observation time. These results suggest that cuteness perception controls physical distance differently depending on the target of evaluation.
History of scholarship and learning. The humanities, Social Sciences
Raghavendra Selvan, Bob Pepin, Christian Igel
et al.
The recent advances in deep learning (DL) have been accelerated by access to large-scale data and compute. These large-scale resources have been used to train progressively larger models which are resource intensive in terms of compute, data, energy, and carbon emissions. These costs are becoming a new type of entry barrier to researchers and practitioners with limited access to resources at such scale, particularly in the Global South. In this work, we take a comprehensive look at the landscape of existing DL models for medical image analysis tasks and demonstrate their usefulness in settings where resources are limited. To account for the resource consumption of DL models, we introduce a novel measure to estimate the performance per resource unit, which we call the PePR score. Using a diverse family of 131 unique DL architectures (spanning 1M to 130M trainable parameters) and three medical image datasets, we capture trends about the performance-resource trade-offs. In applications like medical image analysis, we argue that small-scale, specialized models are better than striving for large-scale models. Furthermore, we show that using existing pretrained models that are fine-tuned on new data can significantly reduce the computational resources and data required compared to training models from scratch. We hope this work will encourage the community to focus on improving AI equity by developing methods and models with smaller resource footprints.
Many researchers confirm that the time we live nowadays is witnessing conflicts and wars which made human being face psychological and biological stressors. Psychological literature suggested that such stressors could affect negatively our psychological and biological well-being. Therefore, so many studies were conducted investigating the negative impact of such stressors on personality and how our defense mechanisms could face such difficulties. The current study aims to identify the effectiveness of a counseling program in reducing anxiety disorders and strengthening the psychological immune system among students. To achieve the aims of the current study, the researchers adopted the experimental method and counseling program. The program was applied among a sample of (80) female students who got the highest scores on the Psychological Immune System Scale (PISS) and Taylor's Explicit Anxiety Scale which consists of 48 items. The group was divided into two groups (experimental and control group). Each group consisted of (40) students. After conducting the parity process in the variables of age, number of family members, birth order, economic and education level, the results showed that the counseling program was effective in increasing the effectiveness of psychological immunity and reducing the manifestations of anxiety among the experimental group. The recommendations and suggestions are discussed.
History of scholarship and learning. The humanities
While reinforcement learning (RL) has become a more popular approach for robotics, designing sufficiently informative reward functions for complex tasks has proven to be extremely difficult due their inability to capture human intent and policy exploitation. Preference based RL algorithms seek to overcome these challenges by directly learning reward functions from human feedback. Unfortunately, prior work either requires an unreasonable number of queries implausible for any human to answer or overly restricts the class of reward functions to guarantee the elicitation of the most informative queries, resulting in models that are insufficiently expressive for realistic robotics tasks. Contrary to most works that focus on query selection to \emph{minimize} the amount of data required for learning reward functions, we take an opposite approach: \emph{expanding} the pool of available data by viewing human-in-the-loop RL through the more flexible lens of multi-task learning. Motivated by the success of meta-learning, we pre-train preference models on prior task data and quickly adapt them for new tasks using only a handful of queries. Empirically, we reduce the amount of online feedback needed to train manipulation policies in Meta-World by 20$\times$, and demonstrate the effectiveness of our method on a real Franka Panda Robot. Moreover, this reduction in query-complexity allows us to train robot policies from actual human users. Videos of our results and code can be found at https://sites.google.com/view/few-shot-preference-rl/home.
Offline or batch reinforcement learning seeks to learn a near-optimal policy using history data without active exploration of the environment. To counter the insufficient coverage and sample scarcity of many offline datasets, the principle of pessimism has been recently introduced to mitigate high bias of the estimated values. While pessimistic variants of model-based algorithms (e.g., value iteration with lower confidence bounds) have been theoretically investigated, their model-free counterparts -- which do not require explicit model estimation -- have not been adequately studied, especially in terms of sample efficiency. To address this inadequacy, we study a pessimistic variant of Q-learning in the context of finite-horizon Markov decision processes, and characterize its sample complexity under the single-policy concentrability assumption which does not require the full coverage of the state-action space. In addition, a variance-reduced pessimistic Q-learning algorithm is proposed to achieve near-optimal sample complexity. Altogether, this work highlights the efficiency of model-free algorithms in offline RL when used in conjunction with pessimism and variance reduction.
Regardless of country and age, the importance of history education is always being emphasized. Although the importance of history education is being emphasized in Korea, there are many difficulties in getting students to understand history properly through school classes alone, and it is also difficult to attract students to participate in classes. The effectiveness of education using games has been proven 20 years ago, and the demand for game-based education is gradually increasing in the current education world, which is becoming more open. In this paper, based on the effects proven through research on the existing game-based education, the improvement of historical thinking ability, experiential history learning, and the problems of game-based education introduced in the ESN report and the discomfort of teachers who participated in the education were improved. A plan was suggested to select and use games suitable for basic education. In this thesis, we selected a history game with a clear historical and periodic background and without distortion of history, and experimented with teaching using games focusing on historical thinking and empirical history learning. The learning achievement of textbook-based education was compared.
The article analyzes the features of the entrepreneurial legal personality of juvenile entrepreneurs and the mechanism of its formation and implementation. The author makes a proposal to indicate in the registration certificate of an entrepreneur that the person is under age. The author analyses particular features of the protection of the legal capacity of juvenile entrepreneurs in business activities.
Law, History of scholarship and learning. The humanities
Temporal-Difference (TD) learning is a general and very useful tool for estimating the value function of a given policy, which in turn is required to find good policies. Generally speaking, TD learning updates states whenever they are visited. When the agent lands in a state, its value can be used to compute the TD-error, which is then propagated to other states. However, it may be interesting, when computing updates, to take into account other information than whether a state is visited or not. For example, some states might be more important than others (such as states which are frequently seen in a successful trajectory). Or, some states might have unreliable value estimates (for example, due to partial observability or lack of data), making their values less desirable as targets. We propose an approach to re-weighting states used in TD updates, both when they are the input and when they provide the target for the update. We prove that our approach converges with linear function approximation and illustrate its desirable empirical behaviour compared to other TD-style methods.
The two main impediments to continual learning are catastrophic forgetting and memory limitations on the storage of data. To cope with these challenges, we propose a novel, cognitively-inspired approach which trains autoencoders with Neural Style Transfer to encode and store images. During training on a new task, reconstructed images from encoded episodes are replayed in order to avoid catastrophic forgetting. The loss function for the reconstructed images is weighted to reduce its effect during classifier training to cope with image degradation. When the system runs out of memory the encoded episodes are converted into centroids and covariance matrices, which are used to generate pseudo-images during classifier training, keeping classifier performance stable while using less memory. Our approach increases classification accuracy by 13-17% over state-of-the-art methods on benchmark datasets, while requiring 78% less storage space.
Various automatic curriculum learning (ACL) methods have been proposed to improve the sample efficiency and final performance of deep reinforcement learning (DRL). They are designed to control how a DRL agent collects data, which is inspired by how humans gradually adapt their learning processes to their capabilities. For example, ACL can be used for subgoal generation, reward shaping, environment generation, or initial state generation. However, prior work only considers curriculum learning following one of the aforementioned predefined paradigms. It is unclear which of these paradigms are complementary, and how the combination of them can be learned from interactions with the environment. Therefore, in this paper, we propose a unified automatic curriculum learning framework to create multi-objective but coherent curricula that are generated by a set of parametric curriculum modules. Each curriculum module is instantiated as a neural network and is responsible for generating a particular curriculum. In order to coordinate those potentially conflicting modules in unified parameter space, we propose a multi-task hyper-net learning framework that uses a single hyper-net to parameterize all those curriculum modules. In addition to existing hand-designed curricula paradigms, we further design a flexible memory mechanism to learn an abstract curriculum, which may otherwise be difficult to design manually. We evaluate our method on a series of robotic manipulation tasks and demonstrate its superiority over other state-of-the-art ACL methods in terms of sample efficiency and final performance.
My response to Susanne Lachenicht’s thought-provoking article is a brief attempt to take up her call to write histories that lead not to absolute certainties but to more understanding of the complexities of the past. I focus on documentation, border control, and citizenship in the Early Roman Empire to illustrate some of the radically different ways these were conceptualized and practiced in a premodern multiethnic empire like Rome than in a contemporary nation-state today. Passports, for example, and border control as we know it, did not exist, and migration was not tied to citizenship status. But the account I offer is deliberately tentative and full of qualifications to emphasize the real methodological challenges the study of this subject poses on account of fragmentary literary and material records and the numerous difficulties of interpreting these. I conclude by pointing out both the benefits and the limitations of framing history as a discipline from which one can learn. On the one hand, understanding how seemingly universal categories such as ‘citizen’ and ‘migrant’ are dynamic and constructed rather than static and natural can nuance public debates in nation-states which receive high numbers of migrants (like Germany, Lachenicht’s starting point) by countering ahistorical narratives of a monolithic and sedentary identity. On the other hand, knowledge of the past does not necessarily lead to moral edification.