Federated recommendation (FedRec) can train personalized recommenders without collecting user data, but the decentralized nature makes it susceptible to poisoning attacks. Most previous studies focus on targeted attack promote certain items, while untargeted that aims degrade overall performance of FedRec system remains less explored. In fact, attacks disrupt experience and bring severe financial loss service provider. However, existing methods are either inapplicable or ineffective against...

As users implicitly express their preferences to items on many real-world applications, the implicit feedback based collaborative filtering has attracted much attention in recent years. Pairwise methods have shown state-of-the-art solutions for dealing with feedback, assumption that prefer observed unobserved items. However, each user, huge are not equal represent her preference. In this paper, we propose a Multiple Ranking (MPR) approach, which relaxes simple pairwise preference previous...

The implicit feedback based collaborative filtering (CF) has attracted much attention in recent years, mainly because users implicitly express their preferences many real-world scenarios. current mainstream pairwise methods optimize the Area Under Curve (AUC) and are empirically proved to be helpful exploit binary relevance data, but lead either not address ranking problem, or specifically focus on top- <inline-formula><tex-math notation="LaTeX">$k$</tex-math></inline-formula>...

Abstract Seed coat mucilage plays an important role in promoting seed germination under adversity. Previous studies have shown that Arabidopsis thaliana MYB52 (AtMYB52) can positively regulate accumulation. However, the of Brassica napus ( BnaMYB52 ) accumulation and tolerance to osmotic stress during remains largely unknown. We cloned BnaA09.MYB52 coding domain sequence from B. cv ZS11, identified its conserved protein domains elucidated relationship with homologues a range plant species....

How to identify high-potential talent (HIPO) earlier in their career always has strategic importance for human resource management. While tremendous efforts have been made this direction, most existing approaches are still based on the subjective selection of experts. This could lead unintentional bias and inconsistencies. To end, paper, we propose a neural network dynamic social profiling approach quantitatively identifying HIPOs from newly-enrolled employees by modeling dynamics behaviors...

Physics-guided machine learning (PGML) has become a prevalent approach in studying scientific systems due to its ability integrate theories for enhancing (ML) models. However, most PGML approaches are tailored isolated and relatively simple tasks, which limits their applicability complex involving multiple interacting processes numerous influencing features. In this paper, we propose \textit{\textbf{P}hysics-\textbf{G}uided \textbf{F}oundation \textbf{M}odel (\textbf{PGFM})} that combines...

Physics-guided machine learning (PGML) has become a prevalent approach in studying scientific systems due to its ability integrate theories for enhancing (ML) models. However, most PGML approaches are tailored isolated and relatively simple tasks, which limits their applicability complex involving multiple interacting processes numerous influencing features. In this paper, we propose Physics-Guided Foundation Model (PGFM) that combines pre-trained ML models physics-based leverages...

Water temperature can vary substantially even across short distances within the same sub-watershed. Accurate prediction of stream water at fine spatial resolutions (i.e., scales, ≤ 1 km) enables precise interventions to maintain quality and protect aquatic habitats. Although spatiotemporal models have made substantial progress in spatially coarse time series modeling, challenges persist predicting scales due lack data that scale. To address problem insufficient fine-scale data, we propose a...

Recent advances in Large Language Models (LLMs) have demonstrated significant potential the field of Recommendation Systems (RSs). Most existing studies focused on converting user behavior logs into textual prompts and leveraging techniques such as prompt tuning to enable LLMs for recommendation tasks. Meanwhile, research interest has recently grown multimodal systems that integrate data from images, text, other sources using modality fusion techniques. This introduces new challenges...

Environmental ecosystems exhibit complex and evolving dynamics over time, making the modeling of non-stationary processes critically important. However, traditional methods often rely on static models trained entire datasets, failing to capture drastically fluctuating characteristics. Dynamically adjusting data is challenging, as they can easily either lag behind new trends or overfit newly received data. To address these challenges, we propose Domain-Adaptive Continual Meta-Learning (DACM)...

Click-Through Rate (CTR) prediction of intelligent marketing systems is great importance, in which feature interaction selection plays a key role. Most approaches model interactions features by the same pre-defined operation under expert guidance, among improper may bring unnecessary noise and complicate training process. To that end, this paper, we aim to adaptively evolve select proper operations interact on pairs task guidance. Inspired natural evolution, propose general Cognitive...

Cognitive diagnosis, a fundamental task in education area, aims at providing an approach to reveal the proficiency level of students on knowledge concepts. Actually, monotonicity is one basic conditions cognitive diagnosis theory, which assumes that student's monotonic with probability giving right response test item. However, few previous methods consider during optimization. To this end, we propose Item Response Ranking framework (IRR), aiming introducing pairwise learning into well model...

Intelligent Financial Advisors (IFAs) in online financial applications (apps) have brought new life to personal investment by providing appropriate and high-quality portfolios for users.In real-world scenarios, identifying potential clients is a crucial issue IFAs, i.e., users who are willing purchase the portfolios.Thus, extracting useful information from various characteristics of further predicting their inclination urgent.However, two critical problems encountered real practice make this...

One-hot encoder accompanied by a softmax loss has become the default configuration to deal with multiclass problem, and is also prevalent in deep learning (DL) based recommender systems (RS). The standard process of such methods fit model outputs one-hot encoding ground truth, referred as hard target. However, it known that these targets largely ignore ambiguity unobserved feedback RS, thus may lead sub-optimal generalization performance. In this work, we propose SoftRec, new RS optimization...

Click-through rate (CTR) prediction is fundamental in many industrial applications, such as online advertising and recommender systems. With the development of platforms, sequential user behaviors grow rapidly, bringing us great opportunity to better understand preferences.However, it extremely challenging for existing models effectively utilize entire behavior history each user. First, there a lot noise long histories, which can seriously hurt performance. Second, feeding sequence directly...

It is always challenging task for students to select right universities. For students, graduate job placement the most important component of university quality. However, existing evaluation methods predominantly depend on either subjective criteria, such as perceived quality learning environment and academic prestige, or factors like faculty excellence, which may not provide a precise indication placement. Indeed, there still lack data-driven approach accurately measure based employment...

Every organization has organizational networks for exchange of ideas and information. It is believed that network analysis (ONA) can help the business be more effective. While considerable research efforts have been made visualizing analyzing relationships in networks, it lacks a holistic way to model complex social structures rich semantic information these networks. Indeed, employee behaviors occur across different communication platforms, such as email instant messaging systems, which...

This paper introduces Alpha-Beta Sampling (ABS) strategy, which is particularly intended for the sampling problem of pairwise ranking in one-class collaborative filtering (PROCCF). Specifically, ABS strategy places more emphasis on such training examples, including positive item with a lower preference score and negative items higher each gradient step. Then, we provide corresponding proofs from both perspectives. First, prove that sampled examples by can update model parameters large...

With the accelerated technology development, technological trend forecasting through patent mining has become a hot issue for high-tech companies. In this term, extensive attention been attracted to knowledge flows (TKF), i.e., predicting directional of from one field another. However, existing studies either rely on labor intensive empirical analysis or do not consider intrinsic characteristics inherent in TKF, including double-faced aspects (i.e., act as both source and target) nodes,...