Recently, the notion of Multifactorial Optimization (MFO) has emerged as a promising approach for evolutionary multi-tasking by automatically exploiting latent synergies between optimization problems, simply through solving them together in an unified representation space [1]. It aims to improve convergence characteristics across multiple problems at once seamlessly transferring knowledge them. In [1], efficacy MFO been studied specific mode transfer form implicit genetic chromosomal...

Dynamic multiobjective optimization problem (DMOP) denotes the problem, which contains objectives that may vary over time. Due to widespread applications of DMOP existed in reality, has attracted much research attention last decade. In this article, we propose solve DMOPs via an autoencoding evolutionary search. particular, for tracking dynamic changes a given DMOP, autoencoder is derived predict moving Pareto-optimal solutions based on nondominated obtained before occurs. This can be easily...

In contrast to traditional recommender systems which usually pay attention users' general and long-term preferences, sequential recommendation (SR) can model dynamic intents based on their behaviour sequences suggest the next item(s) them. However, most of existing models learn ranking score an item only its relevance property, personalized user demands in terms different learning objectives, such as diversity, tail novelty or recency, have been played essential roles multi-objective (MOR),...

Dynamic multiobjective optimization problem (DMOP) denotes the which varies over time. As changes in DMOP may exist some patterns that are predictable, to solve DMOP, a number of research efforts have been made develop evolutionary search with prediction approaches estimate problem. A common practice existing is predict change Pareto-optimal solutions (POS) based on historical obtained decision space. However, occur both and objective spaces. Prediction only space thus not be able give...

Drug discovery is an expensive and risky process. To combat the challenges in drug discovery, increasing number of researchers pharmaceutical companies recognize benefits utilizing computational techniques. Evolutionary computation (EC) offers promise as most problems are essentially complex optimization beyond conventional algorithms. EC methods have been widely applied to solve these especially lead com-pound generation molecular virtual evaluation, substantially speeding up process...

In this paper, we propose a Knowledge-enhanced Hierarchical Attention for community question answering with Multi-task learning and Adaptive (KHAMA). First, hierarchical attention network to fully fuse knowledge from input documents base (KB) by exploiting the semantic compositionality of sequences. The external factual helps recognize background (entity mentions their relationships) eliminate noise information long that have sophisticated syntactic structures. addition, build multiple CQA...

Real Time Strategy (RTS) games require macro strategies as well micro to obtain satisfactory performance since it has large state space, action and hidden information. This paper presents a novel hierarchical reinforcement learning model for mastering Multiplayer Online Battle Arena (MOBA) games, sub-genre of RTS games. The novelty this work are: (1) proposing framework, where agents execute by imitation carry out micromanipulations through learning, (2) developing simple self-learning...

Evolutionary multitasking (EMT) has attracted much attention in the community of evolutionary computation recently. It intends to improve performance optimization on multiple problems via knowledge learning and transfer across them while processes progress online. Existing EMT paradigms can be classified as explicit (EEMT) implicit (IEMT) according mechanisms adopted transfer. With additional modules, EEMT often brings flexible algorithmic designs effective against IEMT. However, most...

Construction project life-cycle management should be based on the visualization of a virtual building, through establishment Building Information Model in phase architectural design as information carrier to realize complete integration. This enables all phases and territories whole building achieve in-time information-sharing so overcome traditional territory pattern. also improves running mode during design, costing, construction operation.

Information leakage in the medical industry has become an urgent problem to be solved field of Internet security. However, due need for automated or semiautomated authorization management privacy protection big data environment, traditional model cannot adapt this complex open environment. Although some scholars have studied risk assessment disclosure it is still initial stage exploration. This paper analyzes key indicators that affect security and leakage, including user access behavior...

In recent years, fake news on social media has become a significant threat to societal security, elevating detection research priority. Among various strategies, fact-checking methods stand out for their accuracy, leveraging evidence from dedicated fact databases. However, these often retrieve raw truth, including vast amounts of irrelevant data, based semantic similarity. This approach results in information redundancy and risks missing the nuanced differences between truth. As result,...

Dynamic multi-objective optimization problem (D-MOP) is widely existed in many real-world applications. Over the years, DMOP has attracted research attentions literature. The adaptive indicator-based evolutionary algorithm (IBEA2) a recently proposed (MOEA). It demonstrated strong search capability on commonly used benchmarks over state-of-the-art MOEAs. However, as adaptation of parameter <tex xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Evolutionary Multi-Tasking (EMT), which solves multiple optimization tasks simultaneously, is a burgeoning topic in the area of evolutionary computation. As EMT transfers useful knowledge across to guide search while process progresses online, superior performance has been obtained many recent attempts. Autoencoding multitasking recently proposed algorithm, employs single-layer denoising auto-encoder for transfer. However, since autoencoding (AEEMT) algorithm learns relationships between...

Solving dynamic multi-objective optimization problem (DMOP) requires optimizing multiple conflicting objectives simultaneously. When a is detected in the changing environment, most of existing prediction-based strategies predict trajectory Pareto-optimal solutions (POS), based on historical obtained solution space. In this paper, we present new prediction method to track moving optima for solving DMOP. contrast approaches, propose build model objective As evaluation DMOP front (POF),...