Artificial intelligence (AI) has surpassed top human players in a variety of games. In imperfect information games, these achievements have primarily been driven by Counterfactual Regret Minimization (CFR) and its variants for computing Nash equilibrium. However, most existing research has focused on maximizing payoff, while largely neglecting the importance of strategic diversity and the need for varied play styles, thereby limiting AI’s adaptability to different user preferences. To address this gap, we propose Preference-CFR (Pref-CFR), a novel method that incorporates two key parameters: preference degree and vulnerability degree. These parameters enable the AI to adjust its strategic distribution within an acceptable performance loss threshold, thereby enhancing its adaptability to a wider range of strategic demands. In our experiments with Texas Hold’em, Pref-CFR successfully trained Aggressive and Loose Passive styles that not only match original CFR-based strategies in performance but also display clearly distinct behavioral patterns. Notably, for certain hand scenarios, Pref-CFR produces strategies that diverge significantly from both conventional expert heuristics and original CFR outputs, potentially offering novel insights for professional players.

Counterfactual Regret Minimization (CFR) and its variants are widely recognized as effective algorithms for solving extensive-form imperfect information games. Recently, many improvements have been focused on enhancing the convergence speed of the CFR algorithm. However, most of these variants are not applicable under Monte Carlo (MC) conditions, making them unsuitable for training in large-scale games. We introduce a new MC-based algorithm for solving extensive-form imperfect information games, called MCCFVFP (Monte Carlo Counterfactual Value-Based Fictitious Play). MCCFVFP combines CFR’s counterfactual value calculations with fictitious play’s best response strategy, leveraging the strengths of fictitious play to gain significant advantages in games with a high proportion of dominated strategies. Experimental results show that MCCFVFP achieved convergence speeds approximately 20\%$\sim$50\% faster than the most advanced MCCFR variants in games like poker and other test games.

In many industrial cooperative design systems, experts from different sub-systems propose one sub-solution or more. One sub-solution of a sub-system may be dependent on sub-solutions from other sub-systems. We need to describe dependencies between sub-solutions, choose one sub-solution from each sub-system respectively, and then combine them into an integrated solution of the system. Normally, we get many integrated solutions. The process of finding integrated solutions is called solving the Cooperative Design Integrated Solution Problem (CDIS Problem) in this paper. We propose a cooperative design model based on vertex-colored graph, which describes the CDIS Problem. In this model, CDIS Problem can be regarded as a problem of finding all simple vertex-colored sub-graphs. To solve the problem, we design a search algorithm by combining Bread-First-Search and trackback. Application of the algorithm to experimental graph data is presented. The results show that this algorithm runs effectively in practice.

In order to solve the dynamic substitution problem of service components in dynamic service environment, the paper proposed a solution, which contains a model called logic-interface orchestration, and a method for service composition based on logic-interface orchestration that realizes the model. The model involves the logic-interfaces that will be bound with service components to compose the service. Each logic-interface has a composition queue of service components to achieve self-compensation. The method uses XML to describe the binding relationships between the logic-interfaces and service components. By updating the XML information, the composite service can adapt to the dynamic changes of service environment. The solution was applied to the Enterprise Service Bus Subsystem. It turned out that the solution facilitates the dynamic replacement of service components and can improve the reliability of the composite service.

In the process of collaborative design, a large number of versions are produced. Storing and accessing versions is very frequent, so version storage in the process of collaborative design has become the core issue of version management. In recent years, how to reduce the storage space and improve the retrieval efficiency has become a research hotspot, however the research of version security is still in its infancy. This paper proposes a model of version storage based on circular linked list, depicts the construction algorithm of the model and analyzes the strategy of version resumption and version security. The model combines the full storage, difference storage and positive difference storage, which reduces the space of version storage, improves the access speed, and ensures that any version can be resumed once lost, which improves the security of the model.

It would be valuable to use metrics to identify the fault-proneness of software modules. However, few research works are on how to select appropriate metrics for fault-proneness prediction currently. We conduct a large-scale comparative experiment of nine different software metrics reduction methods over eleven public-domain data sets from the NASA metrics data repository. The Naive Bayes data miner, with a log-filtering preprocessor on the numeric data, is utilized to construct the prediction model. Comparisons are based on the analysis of variance. Our conclusion is that, reduction methods of software metrics are important to build adaptable and robust software fault-proneness prediction models. Given our results on Naive Bayes and log-filtering, discrete wavelet transformation outperforms other reduction methods, and correlation-based feature selection with genetic search algorithm and information gain can also obtain better predicted performance.