China
Africa
Explore chapters and articles related to this topic
Agents in Economic Markets and Games
Published in Mariam Kiran, X-Machines for Agent-Based Modeling, 2017
Researchers have used the iterated prisoner’s dilemma game to draw important conclusions on behavior of group selection or mutual altruism in real individuals. The gaining of trust among individuals when coming together in groups is often viewed as an evolutionary process which allows evolution of cooperative behaviors. Politics exhibits a PD scenario, illustrating when the country has to make decisions in spending money on its military expansion or reducing weapons. Advertising in economics is viewed as an example of a PD scenario, where firms are competing against each other for sales. They have to decide whether they need to advertise or not depending on whether the other firm has advertised. Their decisions and the times at which they make them would affect their sales.
A review of game theory models of lane changing
Published in Transportmetrica A: Transport Science, 2020
To date, there have been few studies about EGT-based LC models. In general, evolutionary game theory applied in LC models tends to explain drivers' progressive cooperation interactions from the perspective of the whole society. Cortés-Berrueco, Gershenson, and Stephens (2016) assumed that all agents in a constant population are able to arrange themselves to achieve the evolution of cooperation with others, and then put them into a traffic simulation controlled by a probabilistic Cellular Automaton (CA) model named GLAI for updating key information needed. In the experiment, the player who decides the cooperative strategies will pay the cost that the other player receives as a reward. Also, the cooperative probabilities of players will be updated as their behaviors towards ESS (previously mentioned in Section 4) according to one of the five protocols (Kin Selection, Direct Reciprocity, Indirect Reciprocity, Network Reciprocity or Group Selection) presented by Nowak (2006). The results of the simulation test manifest that the human cooperation behavior only occurs in certain values of the density, which shows that the traffic condition is the main factor for cooperative LC behaviors.
Recycling scheduling of urban damaged shared bicycles based on improved genetic algorithm
Published in International Journal of Logistics Research and Applications, 2019
Xuedong Liang, Guangsen Si, Leilei Jiao, Zhi Li
Classical GA selection operations often use fitness ratio, roulette and other methods. Roulette is the most commonly used method. The uncertainty of this selection method is very large. The selection process will miss the excellent individual. A new improved GA is designed to solve the problem based on the principle of natural population evolution and the Pareto Principle of wealth imbalance. The improved algorithm increases the probability that the individuals with high fitness values are selected, and makes up for the shortcomings of the classical GA in the selection operation. The population is divided into two groups according to the fitness value. The 20% individuals with high fitness value in the population form a group with strong survival ability. The remaining individuals form a group with poor viability. The fitness values of the two groups are different. Therefore, different selection probabilities should be set for different groups. The group selection probability with high fitness value is set to 0.8, and the group selection probability with low fitness value is set to 0.2. The improved algorithm extremely follows the evolution mechanism of natural population. In order to further clarify the improvement process of GA, the above steps are represented in Figure 3.