Papers
Topics
Authors
Recent
Search
2000 character limit reached

A study of the Multicriteria decision analysis based on the time-series features and a TOPSIS method proposal for a tensorial approach

Published 21 Oct 2020 in cs.AI | (2010.11720v1)

Abstract: A number of Multiple Criteria Decision Analysis (MCDA) methods have been developed to rank alternatives based on several decision criteria. Usually, MCDA methods deal with the criteria value at the time the decision is made without considering their evolution over time. However, it may be relevant to consider the criteria' time series since providing essential information for decision-making (e.g., an improvement of the criteria). To deal with this issue, we propose a new approach to rank the alternatives based on the criteria time-series features (tendency, variance, etc.). In this novel approach, the data is structured in three dimensions, which require a more complex data structure, as the \textit{tensors}, instead of the classical matrix representation used in MCDA. Consequently, we propose an extension for the TOPSIS method to handle a tensor rather than a matrix. Computational results reveal that it is possible to rank the alternatives from a new perspective by considering meaningful decision-making information.

Summary

  • The paper introduces a novel tensor-based TOPSIS modification for MCDA that integrates time-series analysis to capture evolving decision criteria.
  • It applies the method to rank countries by HDI using dynamic indicators like averages, variability, and trend slopes, revealing significant ranking shifts.
  • The study incorporates SMAA for robustness assessment, enabling decision-makers to gauge sensitivity to time-dependent feature weights.

Overview of Multicriteria Decision Analysis Using Time-Series Features and TOPSIS Tensorial Approach

The paper, "A study of the Multicriteria decision analysis based on the time-series features and a TOPSIS method proposal for a tensorial approach," presents a novel approach to Multicriteria Decision Analysis (MCDA). This work expands traditional MCDA frameworks by incorporating the evolution of decision criteria over time through time-series data and utilizing tensorial data structures rather than relying solely on static criteria values.

Motivation and Methodology

Typically, MCDA methods evaluate alternatives by using a decision matrix where alternatives and criteria are represented by rows and columns, respectively. However, these approaches neglect the potential insights provided by the evolution of criterion over time. This paper addresses this gap by proposing a methodology that maps decision matrices into a higher-dimensional space using time-series features. This approach acknowledges tendencies, variances, and other time-dependent performance features by representing data with tensors.

The authors propose an adaptation of the TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) method to handle tensorial data structures. Classical TOPSIS evaluates alternatives based on their distance from an ideal solution, which is impractical for higher-dimensional tensor structures. Here, the authors' extended TOPSIS method extracts relevant features from time-series data, normalizes and weights these features, and then computes ideal positive and negative points to establish a ranking of alternatives based on dynamic decision criteria.

Numerical Results and Implications

The study applies the proposed methodology to rank countries by Human Development Index (HDI), considering criteria such as life expectancy, education, and income over time. Notably, the approach reveals varied country rankings when different time-series features are emphasized—current data, averages, coefficients of variation, and slope coefficients. For instance, countries like Russia present a prominent shift in ranking when the slope coefficient (indicative of improvement trends) is considered, suggesting a potential underestimation of their long-term performance in traditional static frameworks.

The use of stochastic multicriteria acceptability analysis (SMAA) further provides a sensitivity analysis on ranking fluctuation against changes in feature weights. This robustness analysis helps decision-makers visualize alternative rankings as feature relevance changes, facilitating a comprehensive understanding of potential decision pathways.

Theoretical and Practical Future Directions

The innovation in this framework is substantial, suggesting an impactful shift in how time-series data can be leveraged in decision analysis. Practically, this framework's ability to leverage time-evolution metrics can significantly benefit fields such as economics, public policy, and health sectors, wherein decision criteria naturally evolve. Theoretically, this tensor-based approach opens new avenues for research in high-dimensional decision frameworks and the computational methods efficiently managing these structures. Future advancements might involve enhancing the interpretability of tensorial MCDA outputs and integrating further machine-learning techniques to predict future criterion trajectories.

In summary, incorporating temporal evolution within decision criteria offers a refined lens through which to evaluate alternatives, presenting decision-makers with the opportunity for more informed and potentially more effective decisions grounded in dynamic data analysis.

Paper to Video (Beta)

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Collections

Sign up for free to add this paper to one or more collections.