A Multi-objective Genetic Algorithm for the Software Project Scheduling Problem

Abstract

Abstract. The software project scheduling problem considers the assignment of employees to project tasks with the aim of minimizing the project cost and delivering the project on time. Recent research takes into account that each employee is proficient in some development tasks only, which requiere specific skills. However, this cannot be totally applied in the Mexican context due to software companies do not categorize their employees by software skills, but by their skill level instead. In this study we propose a model that is closer to how software companies operate in Mexico. Moreover, we propose a multi-objective genetic algorithm for solving benchmark instances of this model. Results show that our proposed genetic algorithm performs similarly to two recent approaches and that it finds better multi-objective solutions when they are compared to those found by a well-known multi-objective optimizer.

The Importance of Ethics in Socio-Educational Robotics

“…Owners of influential technology companies appeared there and members of academia and science. These are some of the conclusions they reached [5]:…”

“…Social robotics can be defined as the study of robots capable of interacting and communicating with each other, with human beings and with the environment, within the social and cultural structures in which they operate [2,5].…”

Robust Scheduling of Multi-Skilled Workforce Allocation: Job Rotation Approach

“…The skills of programmers in SPSPs have been discussed in many studies so far. In [42], workers are classified according to four different skill levels, beginner, junior, senior, and expert, with an entry-level employee receiving the lowest salary and an expert earning the highest salary. Additionally, tasks are assigned a required skill level, which corresponds to one of the four mentioned.…”

Application of Genetic Algorithms for Decision-Making in Project Management: A Literature Review

“…From the review, it can be inferred that in all cases the representations of solutions (GA chromosome) and encoding strategies are very project-specific, for example, project costs, project duration and project overtime [43]; task start date [11], [35]; a specific mode for each activity and activity lists [7], [23], [24], [33], [38], [41]; human resources [11], [23], [33] and their priorities [12]. The most common types of chromosome encoding in the reviewed studies are binary code [11], [12], [18], [23], [24], [35], [36] and integer number strings [7], [33], [41].…”

“…The most common types of chromosome encoding in the reviewed studies are binary code [11], [12], [18], [23], [24], [35], [36] and integer number strings [7], [33], [41]. Other types also used in some studies are chromosome encoding types, such as the list of strings [8], graph nodes [38] and encoding method proposed by the authors [43]. In two of the reviewed articles [6], [19], the authors do not suggest a specific chromosome encoding method.…”