The decision-making problem for the case of human-machine analysis of input data is considered. It was noted that the use of human-machine decision support systems allows to reduce time and money costs. A multi-channel automated decision-making system is considered, which can generate real-time decisions based on signals coming to it from different channels. All channels are numeric. Channels can be both software tools of artificial intelligence and competent experts who give conclusions on the researched issue.
Two cases were studied: – the case of making decisions regarding the numerical assessment of an object or phenomenon, when the agreed decision must be numerical; – a case of making decisions regarding the fact of the appearance of an object or phenomenon, when the agreed decision must be logical.
Seven rules have been developed for determining the numerical assessment of an object or phenomenon. The rules allow you to take into account the estimates obtained from different channels and the reliability of these channels. Separate rules take into account ratings received from all channels. There are rules that take into account only the evaluations of those channels whose reliability meets the specified limits. This approach ensures a sufficiently reliable decision, according to the needs of the task and the person making the decision.
Four rules have been developed for the case of decision-making regarding the fact of the appearance of an object or phenomenon. These rules, analyzing the numerical estimates received from the channels, produce a solution from the set {True, False}, which corresponds to cases of occurrence/absence of the phenomenon under investigation. The rules take into account the reliability of the channels and, based on the constructed functional dependence, convert the numerical evaluation into a logical one.
The constructed decision-making scheme in multi-channel human-machine decision support systems makes it possible to arbitrarily increase the number of channels in the system. The use of rules that filter out estimates obtained from channels that are not reliable enough for a specific problem will prevent the dispersion of the estimation result due to a large number of channels.
The choice of rules rests with the decision maker or problem owner.
[1] Dolgikh, S., Mulesa, O. (2021). Collaborative Human-AI Decision-Making Systems. IntSol Workshops, 96–105.
[2] Niese, N. D., & Singer, D. J. (2013). Strategic life cycle decision-making for the management of complex Systems subject to uncertain environmental policy. Ocean engineering, 72, 365–374. https://doi.org/10.1016/j.oceaneng.2013.07.020
[3] Mulesa, O., Snytyuk, V., & Myronyuk, I. (2019). Optimal alternative selection models in a multi-stage decision-making process. EUREKA: Physics and Engineering, (6), 43–50. https://doi.org/10.21303/2461-4262.2019.001005
[4] Webler, T., & Tuler, S. (2021). Four decades of public participation in risk decision making. Risk analysis, 41(3), 503–518. https://doi.org/10.1111/risa.13250
[5] Song, W., & Zhu, J. (2019). A multistage risk decision making method for normal cloud model considering behavior characteristics. Applied Soft Computing, 78, 393–406. https://doi.org/10.1016/j.asoc.2019.02.033
[6] Singh, V., Schiebener, J., Müller, S. M., Liebherr, M., Brand, M., & Buelow, M. T. (2020). Country and sex differences in decision making under uncertainty and risk. Frontiers in Psychology, 11, 486. https://doi.org/10.3389/fpsyg.2020.00486
[7] Yoe, C. (2019). Principles of risk analysis: decision making under uncertainty. CRC press. https://doi.org/10.1201/9780429021121
[8] Timiliotis, J., Blümke, B., Serfözö, P. D., Gilbert, S., Ondré-
sik, M., Türk, E.,… & Eckstein, J. (2022). A novel diagnostic decision support system for medical professionals: prospective feasibility study. JMIR Formative Research, 6(3), 29943. https://doi.org/10.2196/29943
[9] Arena, S., Florian, E., Zennaro, I., Orrù, P. F., & Sgarbossa, F. (2022). A novel decision support system for managing predictive maintenance strategies based on machine learning approaches. Safety science, 146, 105529. https://doi.org/10.1016/j.ssci.2021.105529
[10] Dolgikh, S., Mulesa, O. (2021). Covid-19 epidemiological factor analysis: Identifying principal factors with machine. CEUR Workshop Proceedings, 2833, 114–123.
[11] Kirichenko, L., Vitalii, B., Radivilova, T. (2020). Machine learning classification of multifractional Brownian motion realizations. CEUR Workshop Proceedings, 2608, 980–989. https://doi.org/10.32782/cmis/2608-73
[12] Pasman, H. J., Rogers, W. J., & Behie, S. W. (2022). Selecting a method/tool for risk-based decision making in complex situations. Journal of Loss Prevention in the Process Industries, 74, 104669. https://doi.org/10.1016/j.jlp.2021.104669
[13] Harris, D. (2017). Decision making in aviation. Routledge. https://doi.org/10.4324/9781315095080
[14] Li, G., Yang, Y., Zhang, T., Qu, X., Cao, D., Cheng, B., & Li, K. (2021). Risk assessment based collision avoidance decision-making for autonomous vehicles in multi-scenarios. Transportation research part C: emerging technologies, 122, 102820. https://doi.org/10.1016/j.trc.2020.102820.
[15] Li, S., Sun, Y., & Soergel, D. (2018). Automatic decision support for clinical diagnostic literature using link analysis in a weighted keyword network. Journal of medical systems, 42(2), 1–12. https://doi.org/10.1007/s10916-017-0876-3
[16] Marques, G., Agarwal, D., & de la Torre Díez, I. (2020). Automated medical diagnosis of COVID-19 through EfficientNet convolutional neural network. Applied soft computing, 96, 106691. https://doi.org/10.1016/j.asoc.2020.106691