Algorithmic Decision Making with Python Resources: From multicriteria performance records to decision algorithms via bipolar-valued outranking digraphs

Book published by Springer - 1 (2022)

This book describes Python3 programming resources for implementing decision aiding algorithms in the context of a bipolar-valued outranking approach. These computing resources, made available under the ... [more ▼]

This book describes Python3 programming resources for implementing decision aiding algorithms in the context of a bipolar-valued outranking approach. These computing resources, made available under the name Digraph3, are useful in the field of Algorithmic Decision Theory and more specifically in outranking-based Multiple-Criteria Decision Aiding (MCDA). The first part of the book presents a set of tutorials introducing the Digraph3 collection of Python3 modules and its main objects such as bipolar-valued digraphs and outranking digraphs. The second part illustrates in eight methodological chapters multiple-criteria evaluation models and decision algorithms. These chapters are mostly problem oriented and show how to edit a new multiple-criteria performance tableau, how to build a best choice recommendation, how to compute the winner of an election and how to rank or rate with incommensurable criteria. The book's third part presents three real decision case studies, while the fourth part addresses more advanced topics, such as computing ordinal correlations between bipolar-valued outranking digraphs, computing kernels in bipolar-valued digraphs, testing for confidence or stability of outranking statements when facing uncertain or solely ordinal criteria significance weights, and tempering plurality tyranny effects in social choice problems. The last part is more specifically focused on working with undirected graphs, tree graphs and forests. The closing chapter explores comparability, split, interval and permutation graphs. The book is primarily intended for graduate students in management sciences, computational statistics and operations research. Chapters presenting algorithms for ranking multicriteria performance records will be of computational interest for designers of web recommender systems. Similarly, the relative and absolute quantiles-rating algorithms, discussed and illustrated in several chapters, will be of practical interest for public or private performance auditors. [less ▲]

Tutorials for using the Digraph3 Python software collection

Learning material (2022)

The documentation contains a set of tutorials introducing the main objects like digraphs, outranking digraphs and performance tableaux. There is also a tutorial provided on undirected graphs. Some ... [more ▼]

The documentation contains a set of tutorials introducing the main objects like digraphs, outranking digraphs and performance tableaux. There is also a tutorial provided on undirected graphs. Some tutorials are problem oriented and show how to compute the winner of an election, how to build a best choice recommendation, or how to rank or rate with multiple incommensurable performance criteria. Other tutorials concern more specifically operational aspects of computing maximal independent sets (MISs) and kernels in graphs and digraphs. The tutorial about split, interval and permutation graphs is inspired by Martin Ch. Golumbic’s book on Algorithmic Graph Theory and Perfect Graphs. A recent contribution studies the challenging fair intergroup and intragroup pairing problems. Finally, we provide a tutorial on tree graphs and spanning forests. [less ▲]

Algorithmic Decision Theory: Lecture notes and presentation slides

Learning material (2020)

The objective of this course is to introduce students to ADT, a new interdisciplinary field at the intersection of decision theory, discrete mathematics, theoretical computer science and artificial ... [more ▼]

The objective of this course is to introduce students to ADT, a new interdisciplinary field at the intersection of decision theory, discrete mathematics, theoretical computer science and artificial intelligence. ADT proposes new ideas, approaches and tools for supporting decision making processes in presence of massive databases, combinatorial structures, partial and/or uncertain information, and distributed, possibly inter-operating, decision makers. Such problems arise in several real-world decision making problems such as humanitarian logistics, epidemiology, risk assessment and management, e-government, electronic commerce, and the implementation of recommender systems [less ▲]

Computational Statistics: Lecture notes and presentation slides

Learning material (2018)

The objective of this course is to introduce the students to the R language and environment for statistical computing and graphics (a GNU project). In particular, the course proposes effective data ... [more ▼]

The objective of this course is to introduce the students to the R language and environment for statistical computing and graphics (a GNU project). In particular, the course proposes effective data handling and storage solutions as well as useful operators for calculations on arrays, in particular matrices. A selected collection of intermediate tools for data analysis, graphical facilities for data analysis and display either on-screen or on hard copy will be illustrated from examples of statistical analyses. Finally, the course will by the way acquaint the students with a well-developed, simple and effective programming language which includes conditionals, loops, user-defined recursive functions and input and output facilities. [less ▲]

Algorithmic Decision Theory for solving complex decision problems

Presentation (2017, May 03)

Today's decision makers in fields ranging from engineering to psychology, from medicine to economics and/or homeland security are faced with remarkable new technologies, huge amounts of information to ... [more ▼]

Today's decision makers in fields ranging from engineering to psychology, from medicine to economics and/or homeland security are faced with remarkable new technologies, huge amounts of information to help them in reaching good decisions, and the ability to share information at unprecedented speeds and quantities. These tools and resources should lead to better decisions. Yet, the tools bring with them daunting new problems: the massive amounts of data available are often incomplete, unreliable and/or distributed and there is great uncertainty in them; interoperating/distributed decision makers and decision making devices need to be coordinated; many sources of data need to be fused into a good decision; information sharing under new cooperation/competition arrangements raises security problems. When faced with such issues, there are few highly efficient algorithms available to support decision makers. The objective of Algorithmic Decision Theory (ADT) is to improve the ability of decision makers to perform well when facing these new challenges and problems through the use of methods from theoretical computer science, in particular algorithmic methods. The primary goal of ADT is hence to explore and develop algorithmic approaches for solving decision problems arising in a variety of applications areas. Examples include, but are not limited to: - Computational tractability/intractability of social consensus and multiple criteria compromise functions; - Improvement of decision support and recommender systems; - Development of automatic decision devices including on-line decision procedures; - Robust decision making; - Learning for multi-agent systems and other on-line decision devices. This presentation will focus more specifically on multiple criteria decision aiding methodology, the actual research field of the author. [less ▲]

A sparse outranking digraph model for HPC-ranking of big performance tableaux

Scientific Conference (2016, July 05)

In the context of the ongoing GDRI-Algodec "Algorithmic Decision Theory", supported o.a. by the CNRS (France) and the FNR (Luxembourg), we develop multicriteria ranking HPC algorithms for large sets of ... [more ▼]

In the context of the ongoing GDRI-Algodec "Algorithmic Decision Theory", supported o.a. by the CNRS (France) and the FNR (Luxembourg), we develop multicriteria ranking HPC algorithms for large sets of potential decision alternatives: up to several thousand of alternatives evaluated on multiple incommensurable ordinal performance criteria. By using Python3.5 multiprocessing resources and the Digraph3 multicriteria software library, we could linearly rank without ties on the UL HPC gaia-80 machine with 120 single threaded cores and a CPU memory of 2.3 TB, in about four hours (6h05') a huge set of 1'732'051 decision alternatives evaluated on 13 performance criteria by balancing economic, ecological and societal decision objectives. Data input is, on the one side, a 1'732'051 x 13 performance tableau of size 2.6GB, and on the other side, a theoretical outranking space consisting of three trillions (3 x 1012) of pairwise outranking situations. A "small" set of 1000 decision alternatives, in a similar setting, may thus be ranked typically in about 2 seconds. [less ▲]

Comparisons of Heat Map and IFL Technique to Evaluate the Performance of Commercially Available Cloud Providers

in IEEE (Ed.) 2016 IEEE 9th International Conference on Cloud Computing (2016)

Cloud service providers (CSPs) offer different Ser- vice Level Agreements (SLAs) to the cloud users. Cloud Service Brokers (CSBs) provide multiple sets of alternatives to the cloud users according to ... [more ▼]

Cloud service providers (CSPs) offer different Ser- vice Level Agreements (SLAs) to the cloud users. Cloud Service Brokers (CSBs) provide multiple sets of alternatives to the cloud users according to users requirements. Generally, a CSB considers the service commitments of CSPs rather than the actual quality of CSPs services. To overcome this issue, the broker should verify the service performances while recommending cloud services to the cloud users, using all available data. In this paper, we compare our two approaches to do so: a min-max-min decomposition based on Intuitionistic Fuzzy Logic (IFL) and a Performance Heat Map technique, to evaluate the performance of commercially available cloud providers. While the IFL technique provides simple, total order of the evaluated CSPs, Performance Heat Map provides transparent and explanatory, yet consistent evaluation of service performance of commercially available CSPs. The identified drawbacks of the IFL technique are: 1) It does not return the accurate performance evaluation over multiple decision alternatives due to highly influenced by critical feedback of the evaluators; 2) Overall ranking of the CSPs is not as expected according to the performance measurement. As a result, we recommend to use performance Heat Map for this problem. [less ▲]

On boosting Kohler's ranking-by-choosing rule with a quantiles preordering

Scientific Conference (2015, February 06)

Evaluation and decision models with multiple criteria: Case studies

Book published by Springer (2015)

On confident outrankings with multiple criteria of uncertain significance

Scientific Conference (2014, November 11)