ACM Transactions on Intelligent Systems and Technology最新文献

The Facial Action Coding System (FACS) encodes the action units (AUs) in facial images, which has attracted extensive research attention due to its wide use in facial expression analysis. Many methods that perform well on automatic facial action unit (AU) detection primarily focus on modelling various AU relations between corresponding local muscle areas or mining global attention-aware facial features; however, they neglect the dynamic interactions among local-global features. We argue that encoding AU features just from one perspective may not capture the rich contextual information between regional and global face features, as well as the detailed variability across AUs, because of the diversity in expression and individual characteristics. In this paper, we propose a novel Multi-level Graph Relational Reasoning Network (termed MGRR-Net) for facial AU detection. Each layer of MGRR-Net performs a multi-level (i.e., region-level, pixel-wise and channel-wise level) feature learning. On the one hand, the region-level feature learning from the local face patch features via graph neural network can encode the correlation across different AUs. On the other hand, pixel-wise and channel-wise feature learning via graph attention networks (GAT) enhance the discrimination ability of AU features by adaptively recalibrating feature responses of pixels and channels from global face features. The hierarchical fusion strategy combines features from the three levels with gated fusion cells to improve AU discriminative ability. Extensive experiments on DISFA and BP4D AU datasets show that the proposed approach achieves superior performance than the state-of-the-art methods.

Food computing, as a newly emerging topic, is closely linked to human life through computational methodologies. Meal recommendation, a food-related study about human health, aims to provide users a meal with courses constrained from specific categories (e.g., appetizers, main dishes) that can be enjoyed as a service. Historical interaction data, as important user information, is often used by existing models to learn user preferences. However, if user’s preferences favor less healthy meals, the model will follow that preference and make similar recommendations, potentially negatively impacting the user’s long-term health. This emphasizes the necessity for health-oriented and responsible meal recommendation systems. In this paper, we propose a healthiness-aware and category-wise meal recommendation model called CateRec, which boosts healthiness exposure by using nutritional standards as knowledge to guide the model training. Two fundamental questions are raised and answered: 1) How to evaluate the healthiness of meals? Two well-known nutritional standards from the World Health Organisation and the United Kingdom Food Standards Agency are used to calculate the healthiness score of the meal. 2) How to health-orientedly guide the model training? We construct category-wise personalization partial rankings and category-wise healthiness partial rankings, and theoretically analyze that they meet the necessary properties and assumptions required to be trained by the maximum posterior estimator under Bayesian probability. The data analysis confirms the existence of user preferences leaning towards less healthy meals in two public datasets. A comprehensive experiment demonstrates that our CateRec effectively boosts healthiness exposure in terms of mean healthiness score and ranking exposure, while being comparable to the state-of-the-art model in terms of recommendation accuracy.

Recommendation in settings such as e-recruitment and online dating involves distributing limited opportunities, which differs from recommending practically unlimited goods such as in e-commerce or music recommendation. This setting calls for novel approaches to quantify and enforce fairness. Indeed, typical recommender systems recommend each user their top relevant items, such that desirable items may be recommended simultaneously to more and to less qualified individuals. This is arguably unfair to the latter. Indeed, when they pursue such a desirable recommendation (e.g. by applying for a job), they are unlikely to be successful.

To quantify fairness in such settings, we introduce inferiority: a novel (un)fairness measure that quantifies the competitive disadvantage of a user for their recommended items. Inferiority is complementary to envy: a previously-proposed fairness notion that quantifies the extent to which a user prefers other users’ recommendations over their own. We propose to use both inferiority and envy in combination with an accuracy-related measure called utility: the aggregated relevancy scores of the recommended items. Unfortunately, none of these three measures are differentiable, making it hard to optimize them, and restricting their immediate use to evaluation only. To remedy this, we reformulate them in the context of a probabilistic interpretation of recommender systems, resulting in differentiable versions. We show how these loss functions can be combined in a multi-objective optimization problem that we call FEIR (Fairness through Envy and Inferiority Reduction), used as a post-processing of the scores from any standard recommender system.

Experiments on synthetic and real-world data show that the proposed approach effectively improves the trade-offs between inferiority, envy and utility, compared to the naive recommendation and the state of the art method for the related problem of congestion alleviation in job recommendation. We discuss and enhance the practical impact of our findings on a wide range of real-world recommendation scenarios, and we offer implementations of visualization tools to render the envy and inferiority metrics more accessible.

Reconfigurable robots are deployed for applications demanding area coverage, such as cleaning and inspections. Reconfiguration per context, considering beyond a small set of predefined shapes, is crucial for area coverage performance. However, the existing area coverage methods of reconfigurable robots are not always effective and require improvements for ascertaining the intended goal. Therefore, this paper proposes a novel coverage strategy based on internal rehearsals to improve the area coverage performance of a reconfigurable robot. In this regard, a reconfigurable robot is embodied with the cognitive ability to predict the outcomes of its actions before executing them. A genetic algorithm uses the results of the internal rehearsals to determine a set of the robot’s coverage parameters, including positioning, heading, and reconfiguration, to maximize coverage in an obstacle cluster encountered by the robot. The experimental results confirm that the proposed method can significantly improve the area coverage performance of a reconfigurable robot.

L₂ regularization for weights in neural networks is widely used as a standard training trick. In addition to weights, the use of batch normalization involves an additional trainable parameter γ, which acts as a scaling factor. However, L₂ regularization for γ remains an undiscussed mystery and is applied in different ways depending on the library and practitioner. In this paper, we study whether L₂ regularization for γ is valid. To explore this issue, we consider two approaches: 1) variance control to make the residual network behave like an identity mapping and 2) stable optimization through the improvement of effective learning rate. Through two analyses, we specify the desirable and undesirable γ to apply L₂ regularization and propose four guidelines for managing them. In several experiments, we observed that applying L₂ regularization to applicable γ increased 1%–4% classification accuracy, whereas applying L₂ regularization to inapplicable γ decreased 1%–3% classification accuracy, which is consistent with our four guidelines. Our proposed guidelines were further validated through various tasks and architectures, including variants of residual networks and transformers.

In today’s digital landscape, recommender systems have gained ubiquity as a means of directing users towards personalized products, services, and content. However, despite their widespread adoption and a long track of research, these systems are not immune to shortcomings. A significant challenge faced by recommender systems is the presence of biases, which produces various undesirable effects, prominently the popularity bias. This bias hampers the diversity of recommended items, thus restricting users’ exposure to less popular or niche content. Furthermore, this issue is compounded when multiple stakeholders are considered, requiring the balance of multiple, potentially conflicting objectives.

In this paper, we present a new approach to address a wide range of undesired consequences in recommender systems that involve various stakeholders. Instead of adopting a consequentialist perspective that aims to mitigate the repercussions of a recommendation policy, we propose a deontological approach centered around a minimal set of ethical principles. More precisely, we introduce two distinct principles aimed at avoiding overconfidence in predictions and accurately modeling the genuine interests of users. The proposed approach circumvents the need for defining a multi-objective system, which has been identified as one of the main limitations when developing complex recommenders. Through extensive experimentation, we show the efficacy of our approach in mitigating the adverse impact of the recommender from both user and item perspectives, ultimately enhancing various beyond accuracy metrics. This study underscores the significance of responsible and equitable recommendations and proposes a strategy that can be easily deployed in real-world scenarios.

Personalized clinical decision support systems are increasingly being adopted due to the emergence of data-driven technologies, with this approach now gaining recognition in critical care. The task of incorporating diverse patient conditions and treatment procedures into critical care decision-making can be challenging due to the heterogeneous nature of medical data. Advances in Artificial Intelligence (AI), particularly Reinforcement Learning (RL) techniques, enables the development of personalized treatment strategies for severe illnesses by using a learning agent to recommend optimal policies. In this study, we propose a Deep Reinforcement Learning (DRL) model with a tailored reward function and an LSTM-GRU-derived state representation to formulate optimal treatment policies for vasopressor administration in stabilizing patient physiological states in critical care settings. Using an ICU dataset and the Medical Information Mart for Intensive Care (MIMIC-III) dataset, we focus on patients with Acute Respiratory Distress Syndrome (ARDS) that has led to Sepsis, to derive optimal policies that can prioritize patient recovery over patient survival. Both the DDQN (RepDRL-DDQN) and Dueling DDQN (RepDRL-DDDQN) versions of the DRL model surpass the baseline performance, with the proposed model’s learning agent achieving an optimal learning process across our performance measuring schemes. The robust state representation served as the foundation for enhancing the model’s performance, ultimately providing an optimal treatment policy focused on rapid patient recovery.

Domain knowledge is gradually renovating its attributes to exhibit distinct features in autonomy, propelled by the shift of modern transportation systems (TS) towards autonomous TS (ATS) comprising three progressive generations. Knowledge graph (KG) and its corresponding versions can help depict the evolving TS. Given that KG versions exhibit asymmetry primarily due to variations in evolved knowledge, it is imperative to harmonize the evolved knowledge embodied by the entity across disparate KG versions. Hence, this paper proposes a siamese-based graph convolutional network (GCN) model, namely SiG, to address unresolved issues of low accuracy, efficiency, and effectiveness in aligning asymmetric KGs. SiG can optimize entity alignment in ATS and support the analysis of future-stage ATS development. Such a goal is attained through: a) generating unified KGs to enhance data quality, b) defining graph split to facilitate entire-graph computation, c) enhancing GCN to extract intrinsic features, and d) designing siamese network to train asymmetric KGs. The evaluation results suggest that SiG surpasses other commonly employed models, resulting in average improvements of 23.90% and 37.89% in accuracy and efficiency, respectively. These findings have significant implications for TS evolution analysis and offer a novel perspective for research on complex systems limited by continuously updated knowledge.

A strategy refers to the rules that the agent chooses the available actions to achieve goals. Adopting reasonable strategies is challenging but crucial for an intelligent agent with limited resources working in hazardous, unstructured, and dynamic environments to improve the system’s utility, decrease the overall cost, and increase mission success probability. This paper proposes a novel hierarchical strategy decomposition approach based on Bayesian chaining to separate an intricate policy into several simple sub-policies and organize their relationships as Bayesian strategy networks (BSN). We integrate this approach into the state-of-the-art DRL method – soft actor-critic (SAC), and build the corresponding Bayesian soft actor-critic (BSAC) model by organizing several sub-policies as a joint policy. Our method achieves the state-of-the-art performance on the standard continuous control benchmarks in the OpenAI Gym environment. The results demonstrate that the promising potential of the BSAC method significantly improves training efficiency. Furthermore, we extend the topic to the Multi-Agent systems (MAS), discussing the potential research fields and directions.

Many deep learning works on financial time-series forecasting focus on predicting future prices/returns of individual assets with numerical price-related information for trading, and hence propose models designed for univariate, single task and/or unimodal settings. Forecasting for investment and risk management involves multiple tasks in multivariate settings: forecasts of expected returns and risks of assets in portfolios, and correlations between these assets. As different sources/types of time-series influence future returns, risks and correlations of assets in different ways, it is also important to capture time-series from different modalities. Hence, this paper addresses financial time-series forecasting for investment and risk management in a multivariate, multitask and multimodal setting. Financial time-series forecasting is however challenging due to the low signal-to-noise ratios typical in financial time-series, and as intra-series and inter-series relationships of assets evolve across time. To address these challenges, our proposed Temporal Implicit Multimodal Network (TIME) model learns implicit inter-series relationship networks between assets from multimodal financial time-series at multiple time-steps adaptively. TIME then uses dynamic network and temporal encoding modules to jointly capture such evolving relationships, multimodal financial time-series and temporal representations. Our experiments show that TIME outperforms other state-of-the-art models on multiple forecasting tasks and investment and risk management applications.