From Data to Integrated Risk Management and Smart Living: Mathematical Modelling, Statistical Inference, and Decision Making

智能生活是一种强大的趋势，正迅速出现在寻求优化和改善人类生活条件的先进社会中，特别是通过开发基于大量数据的综合风险管理技术，结合先进的统计模型和推理，以做出更明智和更经济的决策。 我的研究计划的主要科学目标是进一步推进对复杂系统的行为、可靠性和弹性进行建模和预测的尖端方法，并培训高素质的人员，这些复杂系统包括经济系统(例如财富和收入在地理区域之间的重新分配)、工业系统(例如可靠性、恢复力、生存能力)、金融和保险(例如风险度量和管理、趋势建模、预测、成本效益分析)。这些领域已经主导了我的研究，并将继续主导我的研究。 具体地说，我的研究在以下几个相互交织的方向上取得了进展： 量化和定性风险的汇总和分配技术，这在风险计量和管理以及决策中至关重要。事实上，许多风险(例如，运营风险、战略风险)源自复杂系统(例如，合作、竞争、模糊、多状态等)。并且可以以定性的形式观察到，而这些形式往往很难有意义地量化。 在使用和不使用足够大的数据集的情况下进行风险回报和成本收益分析。在“有”的情况下，我将致力于开发适当的参数、半参数和非参数的统计推断技术，而在“没有”的情况下，我将关键地依靠基于知识的概率(例如行为分析)、校准和更新技术(例如信息流)。 通过结合随机过程(例如经验和分位数过程、非线性时间序列)的技术以及快速发展的定量心理启发式(例如消费者行为分析)领域的技术来进行风险-回报预测和决策。后一种技术使决策者能够相对快速地评估决策精度，并在现实生活中施加的时间限制之间取得平衡。 系统风险建模在开发静态和动态环境下的风险-回报聚合和分配方法、投资组合多样化和优化技术(例如随机优势)方面发挥着关键作用。 所有这些研究方向都是当今快速发展的智能生活综合风险管理方法的组成部分，该方法一直主导着国内外的研究。精通尖端数学和统计建模、计算和统计推理以及决策的研究人员对智能生活有很大贡献。

Smart Living is a powerful trend that is rapidly emerging in advanced societies that seek to optimize and improve human living conditions by, in particular, developing integrated risk management techniques based on massive data in combination with advanced statistical modelling and inference for making wiser and more economical decisions. The key scientific objective of my research program is to further advance the cutting-edge methodology of, and train highly qualified personnel in, modeling and forecasting the behavior, reliability and resilience of complex systems, such as economic (e.g. re-distribution of wealth and incomes among geographic regions), industrial (e.g. reliability, resilience, survivability), financial and insurance (e.g. risk measurement and management, modelling of trends, forecasting, cost-benefit analysis). These are the areas that have dominated, and will continue dominating, my research. Specifically, my research has been progressing in the following intertwined directions: Aggregation and allocation techniques for quantitative and qualitative risks, which are pivotal in risk measurement and management, as well as in decision making. Indeed, a number of risks (e.g. operational, strategic) originate from complex systems (e.g. co-operative, competing, vague, multi-state, etc.) and are observable in qualitative forms that are frequently difficult to meaningfully quantify. Risk-reward and cost-benefit analyses with and without sufficiently large data sets. Whereas in the “with” case I will work on developing appropriate parametric, semi- and non-parametric techniques of statistical inference, in the “without” case I will crucially rely on knowledge-based probabilities (e.g. behavioral analysis), calibration and updating techniques (e.g. information flows). Risk-reward forecasting and decision-making by combining techniques of stochastic processes (e.g. empirical and quantile processes, non-linear time series) as well as those of the rapidly developing area of quantitative psychological heuristics (e.g. consumer behavioral analysis). The latter techniques allow decision-makers to relatively quickly assess and balance decision-precision with real-life imposed time constraints. Systemic-risk modeling, which plays a pivotal role in developing risk-reward aggregation and allocation methods, portfolio diversification and optimization techniques (e.g. stochastic dominance) in static and dynamic environments. All of these research directions make up an integral part of the nowadays rapidly developing integrated risk management approach to Smart Living that has been dominating research nationally and internationally. Researchers who are well versed in cutting-edge mathematical and statistical modelling, computing and statistical inference, as well as in decision making have much to contribute to Smart Living.