'Bayesian Optimisation for Nurse Scheduling'

A Bayesian optimisation algorithm for a nurse scheduling problem is presented, which involves choosing a suitable scheduling rule from a set for each nurse's assignment. When a human scheduler works, he normally builds a schedule systematically following a set of rules. After much practice, the scheduler gradually masters the knowledge of which solution parts go well with others. He can identify good parts and is aware of the solution quality even if the scheduling process is not yet completed, thus having the ability to finish a schedule by using flexible, rather than fixed, rules. In this paper, we design a more human-like scheduling algorithm, by using a Bayesian optimisation algorithm to implement explicit learning from past solutions. A nurse scheduling problem from a UK hospital is used for testing. Unlike our previous work that used Genetic Algorithms to implement implicit learning [1], the learning in the proposed algorithm is explicit, i.e. we identify and mix building blocks directly. The Bayesian optimisation algorithm is applied to implement such explicit learning by building a Bayesian network of the joint distribution of solutions. The conditional probability of each variable in the network is computed according to an initial set of promising solutions. Subsequently, each new instance for each variable is generated by using the corresponding conditional probabilities, until all variables have been generated, i.e. in our case, new rule strings have been obtained. Sets of rule strings are generated in this way, some of which will replace previous strings based on fitness. If stopping conditions are not met, the conditional probabilities for all nodes in the Bayesian network are updated again using the current set of promising rule strings. For clarity, consider the following toy example of scheduling five nurses with two rules (1: random allocation, 2: allocate nurse to low-cost shifts). In the beginning of the search, the probabilities of choosing rule 1 or 2 for each nurse is equal, i.e. 50%. After a few iterations, due to the selection pressure and reinforcement learning, we experience two solution pathways: Because pure low-cost or random allocation produces low quality solutions, either rule 1 is used for the first 2-3 nurses and rule 2 on remainder or vice versa. In essence, Bayesian network learns 'use rule 2 after 2-3x using rule 1' or vice versa. It should be noted that for our and most other scheduling problems, the structure of the network model is known and all variables are fully observed. In this case, the goal of learning is to find the rule values that maximize the likelihood of the training data. Thus, learning can amount to 'counting' in the case of multinomial distributions. For our problem, we use our rules: Random, Cheapest Cost, Best Cover and Balance of Cost and Cover. In more detail, the steps of our Bayesian optimisation algorithm for nurse scheduling are: 1. Set t = 0, and generate an initial population P(0) at random; 2. Use roulette-wheel selection to choose a set of promising rule strings S(t) from P(t); 3. Compute conditional probabilities of each node according to this set of promising solutions; 4. Assign each nurse using roulette-wheel selection based on the rules' conditional probabilities. A set of new rule strings O(t) will be generated in this way; 5. Create a new population P(t+1) by replacing some rule strings from P(t) with O(t), and set t = t+1; 6. If the termination conditions are not met (we use 2000 generations), go to step 2. Computational results from 52 real data instances demonstrate the success of this approach. They also suggest that the learning mechanism in the proposed approach might be suitable for other scheduling problems. Another direction for further research is to see if there is a good constructing sequence for individual data instances, given a fixed nurse scheduling order. If so, the good patterns could be recognized and then extracted as new domain knowledge. Thus, by using this extracted knowledge, we can assign specific rules to the corresponding nurses beforehand, and only schedule the remaining nurses with all available rules, making it possible to reduce the solution space. Acknowledgements The work was funded by the UK Government's major funding agency, Engineering and Physical Sciences Research Council (EPSRC), under grand GR/R92899/01. References [1] Aickelin U, "An Indirect Genetic Algorithm for Set Covering Problems", Journal of the Operational Research Society, 53(10): 1118-1126,

'Explicit Learning: an Effort towards Human Scheduling Algorithms'

Abstract Scheduling problems are generally NP-hard combinatorial problems, and a lot of research has been done to solve these problems heuristically. However, most of the previous approaches are problem-specific and research into the development of a general scheduling algorithm is still in its infancy. Mimicking the natural evolutionary process of the survival of the fittest, Genetic Algorithms (GAs) have attracted much attention in solving difficult scheduling problems in recent years. Some obstacles exist when using GAs: there is no canonical mechanism to deal with constraints, which are commonly met in most real-world scheduling problems, and small changes to a solution are difficult. To overcome both difficulties, indirect approaches have been presented (in [1] and [2]) for nurse scheduling and driver scheduling, where GAs are used by mapping the solution space, and separate decoding routines then build solutions to the original problem. In our previous indirect GAs, learning is implicit and is restricted to the efficient adjustment of weights for a set of rules that are used to construct schedules. The major limitation of those approaches is that they learn in a non-human way: like most existing construction algorithms, once the best weight combination is found, the rules used in the construction process are fixed at each iteration. However, normally a long sequence of moves is needed to construct a schedule and using fixed rules at each move is thus unreasonable and not coherent with human learning processes. When a human scheduler is working, he normally builds a schedule step by step following a set of rules. After much practice, the scheduler gradually masters the knowledge of which solution parts go well with others. He can identify good parts and is aware of the solution quality even if the scheduling process is not completed yet, thus having the ability to finish a schedule by using flexible, rather than fixed, rules. In this research we intend to design more human-like scheduling algorithms, by using ideas derived from Bayesian Optimization Algorithms (BOA) and Learning Classifier Systems (LCS) to implement explicit learning from past solutions. BOA can be applied to learn to identify good partial solutions and to complete them by building a Bayesian network of the joint distribution of solutions [3]. A Bayesian network is a directed acyclic graph with each node corresponding to one variable, and each variable corresponding to individual rule by which a schedule will be constructed step by step. The conditional probabilities are computed according to an initial set of promising solutions. Subsequently, each new instance for each node is generated by using the corresponding conditional probabilities, until values for all nodes have been generated. Another set of rule strings will be generated in this way, some of which will replace previous strings based on fitness selection. If stopping conditions are not met, the Bayesian network is updated again using the current set of good rule strings. The algorithm thereby tries to explicitly identify and mix promising building blocks. It should be noted that for most scheduling problems the structure of the network model is known and all the variables are fully observed. In this case, the goal of learning is to find the rule values that maximize the likelihood of the training data. Thus learning can amount to 'counting' in the case of multinomial distributions. In the LCS approach, each rule has its strength showing its current usefulness in the system, and this strength is constantly assessed [4]. To implement sophisticated learning based on previous solutions, an improved LCS-based algorithm is designed, which consists of the following three steps. The initialization step is to assign each rule at each stage a constant initial strength. Then rules are selected by using the Roulette Wheel strategy. The next step is to reinforce the strengths of the rules used in the previous solution, keeping the strength of unused rules unchanged. The selection step is to select fitter rules for the next generation. It is envisaged that the LCS part of the algorithm will be used as a hill climber to the BOA algorithm. This is exciting and ambitious research, which might provide the stepping-stone for a new class of scheduling algorithms. Data sets from nurse scheduling and mall problems will be used as test-beds. It is envisaged that once the concept has been proven successful, it will be implemented into general scheduling algorithms. It is also hoped that this research will give some preliminary answers about how to include human-like learning into scheduling algorithms and may therefore be of interest to researchers and practitioners in areas of scheduling and evolutionary computation. References 1. Aickelin, U. and Dowsland, K. (2003) 'Indirect Genetic Algorithm for a Nurse Scheduling Problem', Computer & Operational Research (in print). 2. Li, J. and Kwan, R.S.K. (2003), 'Fuzzy Genetic Algorithm for Driver Scheduling', European Journal of Operational Research 147(2): 334-344. 3. Pelikan, M., Goldberg, D. and Cantu-Paz, E. (1999) 'BOA: The Bayesian Optimization Algorithm', IlliGAL Report No 99003, University of Illinois. 4. Wilson, S. (1994) 'ZCS: A Zeroth-level Classifier System', Evolutionary Computation 2(1), pp 1-18.

Partnering Strategies for Fitness Evaluation in a Pyramidal Evolutionary Algorithm

This paper combines the idea of a hierarchical distributed genetic algorithm with different inter-agent partnering strategies. Cascading clusters of sub-populations are built from bottom up, with higher-level sub-populations optimising larger parts of the problem. Hence higher-level sub-populations search a larger search space with a lower resolution whilst lower-level sub-populations search a smaller search space with a higher resolution. The effects of different partner selection schemes for (sub-)fitness evaluation purposes are examined for two multiple-choice optimisation problems. It is shown that random partnering strategies perform best by providing better sampling and more diversity.

A Pyramidal Evolutionary Algorithm with Different Inter-Agent Partnering Strategies for Scheduling Problems

This paper combines the idea of a hierarchical distributed genetic algorithm with different inter-agent partnering strategies. Cascading clusters of sub-populations are built from bottom up, with higher-level sub-populations optimising larger parts of the problem. Hence higher-level sub-populations search a larger search space with a lower resolution whilst lower-level sub-populations search a smaller search space with a higher resolution. The effects of different partner selection schemes amongst the agents on solution quality are examined for two multiple-choice optimisation problems. It is shown that partnering strategies that exploit problem-specific knowledge are superior and can counter inappropriate (sub-) fitness measurements.

An idiotypic immune network as a short-term learning architecture for mobile robots

A combined Short-Term Learning (STL) and Long-Term Learning (LTL) approach to solving mobile robot navigation problems is presented and tested in both real and simulated environments. The LTL consists of rapid simulations that use a Genetic Algorithm to derive diverse sets of behaviours. These sets are then transferred to an idiotypic Artificial Immune System (AIS), which forms the STL phase, and the system is said to be seeded. The combined LTL-STL approach is compared with using STL only, and with using a handdesigned controller. In addition, the STL phase is tested when the idiotypic mechanism is turned off. The results provide substantial evidence that the best option is the seeded idiotypic system, i.e. the architecture that merges LTL with an idiotypic AIS for the STL. They also show that structurally different environments can be used for the two phases without compromising transferability.

A study on the deployment of GA in a grid computing framework

Dissertação de Mestrado, Engenharia Informática, Faculdade de Ciências e Tecnologia, Universidade do Algarve, 2015

A Comparative Analysis Between Genetic Algorithms and Complex Nonlinear Least Squares on Electrical Impedance Characterization

This paper compares the performance of the complex nonlinear least squares algorithm implemented in the LEVM/LEVMW software with the performance of a genetic algorithm in the characterization of an electrical impedance of known topology. The effect of the number of measured frequency points and of measurement uncertainty on the estimation of circuit parameters is presented. The analysis is performed on the equivalent circuit impedance of a humidity sensor.

A generative design system based on evolutionary and mathematical functions

Previous work by Professor John Frazer on Evolutionary Architecture provides a basis for the development of a system evolving architectural envelopes in a generic and abstract manner. Recent research by the authors has focused on the implementation of a virtual environment for the automatic generation and exploration of complex forms and architectural envelopes based on solid modelling techniques and the integration of evolutionary algorithms, enhanced computational and mathematical models. Abstract data types are introduced for genotypes in a genetic algorithm order to develop complex models using generative and evolutionary computing techniques. Multi-objective optimisation techniques are employed for defining the fitness function in the evaluation process.

Generative and evolutionary models for design

The paper describes three design models that make use of generative and evolutionary systems. The models describe overall design methods and processes. Each model defines a set of tasks to be performed by the design team, and in each case one of the tasks requires a generative or evolutionary design system. The architectures of these systems are also broadly described.

Robot path planning in dynamic environments using a simulated annealing based approach

Mobile robots are widely used in many industrial fields. Research on path planning for mobile robots is one of the most important aspects in mobile robots research. Path planning for a mobile robot is to find a collision-free route, through the robot’s environment with obstacles, from a specified start location to a desired goal destination while satisfying certain optimization criteria. Most of the existing path planning methods, such as the visibility graph, the cell decomposition, and the potential field are designed with the focus on static environments, in which there are only stationary obstacles. However, in practical systems such as Marine Science Research, Robots in Mining Industry, and RoboCup games, robots usually face dynamic environments, in which both moving and stationary obstacles exist. Because of the complexity of the dynamic environments, research on path planning in the environments with dynamic obstacles is limited. Limited numbers of papers have been published in this area in comparison with hundreds of reports on path planning in stationary environments in the open literature. Recently, a genetic algorithm based approach has been introduced to plan the optimal path for a mobile robot in a dynamic environment with moving obstacles. However, with the increase of the number of the obstacles in the environment, and the changes of the moving speed and direction of the robot and obstacles, the size of the problem to be solved increases sharply. Consequently, the performance of the genetic algorithm based approach deteriorates significantly. This motivates the research of this work. This research develops and implements a simulated annealing algorithm based approach to find the optimal path for a mobile robot in a dynamic environment with moving obstacles. The simulated annealing algorithm is an optimization algorithm similar to the genetic algorithm in principle. However, our investigation and simulations have indicated that the simulated annealing algorithm based approach is simpler and easier to implement. Its performance is also shown to be superior to that of the genetic algorithm based approach in both online and offline processing times as well as in obtaining the optimal solution for path planning of the robot in the dynamic environment. The first step of many path planning methods is to search an initial feasible path for the robot. A commonly used method for searching the initial path is to randomly pick up some vertices of the obstacles in the search space. This is time consuming in both static and dynamic path planning, and has an important impact on the efficiency of the dynamic path planning. This research proposes a heuristic method to search the feasible initial path efficiently. Then, the heuristic method is incorporated into the proposed simulated annealing algorithm based approach for dynamic robot path planning. Simulation experiments have shown that with the incorporation of the heuristic method, the developed simulated annealing algorithm based approach requires much shorter processing time to get the optimal solutions in the dynamic path planning problem. Furthermore, the quality of the solution, as characterized by the length of the planned path, is also improved with the incorporated heuristic method in the simulated annealing based approach for both online and offline path planning.

Optimal allocation and sizing of DGs in distribution networks

In this paper, the placement and sizing of Distributed Generators (DG) in distribution networks are determined optimally. The objective is to minimize the loss and to improve the reliability. The constraints are the bus voltage, feeder current and the reactive power flowing back to the source side. The placement and size of DGs are optimized using a combination of Discrete Particle Swarm Optimization (DPSO) and Genetic Algorithm (GA). This increases the diversity of the optimizing variables in DPSO not to be stuck in the local minima. To evaluate the proposed algorithm, the semi-urban 37-bus distribution system connected at bus 2 of the Roy Billinton Test System (RBTS), which is located at the secondary side of a 33/11 kV distribution substation, is used. The results finally illustrate the efficiency of the proposed method.

Load–frequency control : a GA-based multi-agent reinforcement learning

The load–frequency control (LFC) problem has been one of the major subjects in a power system. In practice, LFC systems use proportional–integral (PI) controllers. However since these controllers are designed using a linear model, the non-linearities of the system are not accounted for and they are incapable of gaining good dynamical performance for a wide range of operating conditions in a multi-area power system. A strategy for solving this problem because of the distributed nature of a multi-area power system is presented by using a multi-agent reinforcement learning (MARL) approach. It consists of two agents in each power area; the estimator agent provides the area control error (ACE) signal based on the frequency bias estimation and the controller agent uses reinforcement learning to control the power system in which genetic algorithm optimisation is used to tune its parameters. This method does not depend on any knowledge of the system and it admits considerable flexibility in defining the control objective. Also, by finding the ACE signal based on the frequency bias estimation the LFC performance is improved and by using the MARL parallel, computation is realised, leading to a high degree of scalability. Here, to illustrate the accuracy of the proposed approach, a three-area power system example is given with two scenarios.

Evolutionary algorithms for sustainable building design

This approach to sustainable design explores the possibility of creating an architectural design process which can iteratively produce optimised and sustainable design solutions. Driven by an evolution process based on genetic algorithms, the system allows the designer to “design the building design generator” rather than to “designs the building”. The design concept is abstracted into a digital design schema, which allows transfer of the human creative vision into the rational language of a computer. The schema is then elaborated into the use of genetic algorithms to evolve innovative, performative and sustainable design solutions. The prioritisation of the project’s constraints and the subsequent design solutions synthesised during design generation are expected to resolve most of the major conflicts in the evaluation and optimisation phases. Mosques are used as the example building typology to ground the research activity. The spatial organisations of various mosque typologies are graphically represented by adjacency constraints between spaces. Each configuration is represented by a planar graph which is then translated into a non-orthogonal dual graph and fed into the genetic algorithm system with fixed constraints and expected performance criteria set to govern evolution. The resultant Hierarchical Evolutionary Algorithmic Design System is developed by linking the evaluation process with environmental assessment tools to rank the candidate designs. The proposed system generates the concept, the seed, and the schema, and has environmental performance as one of the main criteria in driving optimisation.

Biologically inspired joint control for a humanoid robot

The GuRoo is a 1.2 m tall, 23 degree of freedom humanoid constructed at the University of Queensland for research into humanoid robotics. The key challenge being addressed by the GuRoo project is the development of appropriate learning strategies for control and coordination of the robot's many joints. The development of learning strategies is seen as a way to side-step the inherent intricacy of modeling a multi-DOF biped robot. This paper outlines the approach taken to generate an appropriate control scheme for the joints of the GuRoo. The paper demonstrates the determination of local feedback control parameters using a genetic algorithm. The feedback loop is then augmented by a predictive modulator that learns a form of feed-forward control to overcome the irregular loads experienced at each joint during the gait cycle. The predictive modulator is based on the CMAC architecture. Results from tests on the GuRoo platform show that both systems provide improvements in stability and tracking of joint control.