Guest editorial

Citation

Jaramillo, J.R. (2022), "Guest editorial", Journal of Modelling in Management, Vol. 17 No. 1, pp. 1-3. https://doi.org/10.1108/JM2-02-2022-324

Analytical modeling and applications

The broader availability of data and the increase of computational power is allowing organizations to find new and valuable insights, giving life to the field of analytics. The field of analytics encompasses the handling and processing of data, from the acquiring of the data to the driving courses of action. Most of the fuzz today is given to the data component of analytics. In fact, some call the field data analytics. But data is not the only component nor the most important one. The other component is the mathematical models used to process the data. In fact, modeling is a key component of the analytics field. There cannot be analytics without models, but there can be analytics with small data. With that in mind, the motivation of this JM2 special issue is to recognize the importance of modeling in the analytics field.

The structure of this special issue is based on the well-known three dimensions of analytics, namely, descriptive, predictive and prescriptive. Descriptive analytics describes current states; predictive analytics looks to what will happen next; and prescriptive analytics emphasizes in finding new and better courses of action. Next, we present 16 articles covering theoretical and applied analytical modeling that covers the three dimensions of analytics with applications in multiple sectors, with an international flavor, and with some contributions to the body theory. The issue starts with five papers in descriptive analytics, then continues with four articles in prescriptive analytics and finishes with seven works in predictive analytics.

Descriptive analytics

Descriptive analytics helps organizations to have a better understanding of their current environment. This section includes a new methodology to manage literature reviews, a novel data envelopment analysis that handles discrete outputs, a study on innovation capability enablers, a technique that handles non-discretionary factors and a study on household shopping behavior in Turkey. Below are the descriptive analytics works: “A methodology for structured literature network meta-analysis” presents a novel way to structure literature components as a network. The network can be represented as a graph, allowing the user to identify important relationships among literature entries, taking advantage of graphic theory metrics and algorithms. “Data envelopment analysis using the binary-data” implements an algorithm that is able to handle binary outputs and to creating a point-to-point frontier that allows to identify benchmarks in a discrete environment. The authors apply the novel model to a health-care study involving stroke. The model allows identifying the most influential risk factors that lead to stroke. This application can be used in multiple fields. “Modelling the continuous innovation capability enablers in Indonesia’s manufacturing industry” presents a fuzzy Delphi model including 21 continuous innovation capability enablers for the country’s industrial companies. The model identifies the level of importance of each enabler, knowing these enablers help manufactures to identify their strengths and weaknesses of their innovation strategies. “Performance measurement and resource sharing among business sub-units in the presence of non-discretionary factors,” this work introduces a data envelopment analysis model that includes non-discretionary inputs because of output partial interactions. The model helps to identify bottleneck points that can benefit from non-discretionary inputs, and it is illustrated with a real case including a set of 17 road maintenance crews. “Determinants of Online Shopping Attitudes of Households in Turkey” applies machine learning techniques to determine the main factors that affect online shopping tendencies in Turkey. The model results provide valuable information to understand the shopping behavior of households.

Predictive analytics

Predictive analytics allows organizations to foresee what is going to happen next. This section contains supply chain application focused on bullwhip impact, an improved electricity demand forecasting, a risk analysis for the Brazilian Navy and ends with an industrial piracy game theory model. Next is a short description of the predictive analytics works: “Operations-Based Classification of Bullwhip Effect” models the bullwhip effect when the variance in production is larger than the variance in demand, using neural networks. The results of the study identify critical operational variables according to the respective industrial sector. The model compares favorably against other machine learning algorithms. The use of the model is illustrated with ten industrial sectors in the county of India. “Electricity demand forecasting using fuzzy hybrid intelligence-based seasonal models” presents a novel forecasting extended fuzzy seasonal method. The proposed method overcomes the limitations of traditional neural networks dealing with uncertainty and seasonality. The method simultaneously models seasonal and fuzzy patterns and structures, as well as the regular non-seasonal patterns. “A Fuzzy approach to assess outsourcing risks in Brazilian Navy Industrial Military Organizations” uses fuzzy analytical hierarchy process to generate information to decision-making in risk management process. In particular, the methodology is useful to deal with inaccurate judgment and different perceptions. Moreover, the model assigns risk levels to each model input. “Analysis of Piracy Trend in ODM Supply Chain” studies the dilemma occurring in the original design manufacturing supply chain in which the foundry is inclined to sell pirate versions of the original product. The analysis is made using a unique five decision model approach. The model helps to identify the conditions in which the foundry could be “inclined” to breach a contract because of financial motivations.

Prescriptive analytics

Prescriptive analytics look for the best course of action. The traditional operations research optimization techniques are part of the prescriptive dimension. This section harbors seven works: a model to optimize bike-sharing services, a dairy products supply chain optimization, a reverse logistics model, an expanded cost-benefit analysis, an optimization model for environmentally friendly energy production and a novel project selection tool and a game theory-based optimization model for infectious waste. Following is a short summary of the prescriptive analytics papers: “Simulation analysis of initial inventory in Bike Sharing Systems” provides a simulation model that optimizes bicycle inventories at multiple parking docks based on bicycle demand history. The objective of the model is to maintain a predefined service level while maintaining bicycle inventories at the most efficient levels. “Production Planning Decision of a Dairy under Supply Disruption” develops a stochastic mixed-integer linear program that optimizes the product mix of dairy products such that profits are maximized. The model also provides insights on the factors that has large effects in the production system. “A robust fuzzy optimization approach for reverse logistics network design with buyback offers” explore the relationship between buyback policies and used products. Buy back policies are a function of quality and age of used products at the time of return among other uncertain characteristics. The problem is modeled using fuzzy mathematical methods, flexible constraint programming and robust fuzzy optimization. “An approach to cost-benefit analysis by competitive advantages with stochastic data” focuses on analyzing the values of alternative’s factors in a stochastic cost-benefit analysis. The problem is addressed using a data envelopment analysis with competitive advantages that evaluates multiple alternatives, leading to the selection of the best one. “A mathematical modeling for energy generation from municipal waste: A case study” presents an optimization solution for the integrated problem of collecting municipal waste (vehicle routing problem with time windows) and the energy generation. The model is a multi-objective problem considering recycling income, energy generation and pollutant emissions. “A Model to Reduce the Risk of Project Selection utilizing Data envelopment analysis” focuses on the selection of projects accounting risk associated to uncertainty. The emphasis is in the risk of choosing a set of projects so that can make a balance in investment versus on collective benefit, taking in account the cost of failures. The problem is modeled as a generalization of the project selection model with limited resources under uncertainty involving negative data. “A novel cooperative model in the collection of infectious waste in COVID-19 pandemic” presents a mathematical programming model with the concept of collaborative game theory to minimize the cost. The solution approach includes a set of methods on fair allocation of cost savings. The model results showcase the advantage of cooperation among contractors. The model is illustrated with a real case in the city of Tehran.