Self organizing optimization and phase transition in reinforcement learning minority game system

Whether the complex game system composed of a large number of artificial intelligence(AI)agents empowered with reinforcement learning can produce extremely favorable collective behaviors just through the way of agent self-exploration is a matter of practical importance.In this paper,we address this question by combining the typical theoretical model of resource allocation system,the minority game model,with reinforcement learning.Each individual participating in the game is set to have a certain degree of intelligence based on reinforcement learning algorithm.In particular,we demonstrate that as AI agents gradually becomes familiar with the unknown environment and tries to provide optimal actions to maximize payoff,the whole system continues to approach the optimal state under certain parameter combinations,herding is effectively suppressed by an oscillating collective behavior which is a self-organizing pattern without any external interference.An interesting phenomenon is that a first-order phase transition is revealed based on some numerical results in our multi-agents system with reinforcement learning.In order to further understand the dynamic behavior of agent learning,we define and analyze the conversion path of belief mode,and find that the self-organizing condensation of belief modes appeared for the given trial and error rates in the AI system.Finally,we provide a detection method for period-two oscillation collective pattern emergence based on the Kullback–Leibler divergence and give the parameter position where the period-two appears.

Genotype and Planting Density Effects on Rooting Traits and Yield in Cotton （Gossypium hirsutum L,）

Root density distribution of plants is a major Indicator of competition between plants and determines resource capture from the solh This experiment was conducted in 2005 at Anyang, located in the Yellow River region, Henan Province, China. Three cotton （Gossyplum hlrsutum L.） cultivars were chosen： hybrid Btcultlvar CRI46, conventional Btcultlvars CRI44 and CRI45. Six planting densities were designed, ranging from 1.5 to 12.0 plants/m^2. Root parameters such as surface area, diameter and length were analyzed by using the DT-SCAN Image analysis method. The root length density （RLD）, root average diameter and root area Index （RAI）, root surface area per unit land area, were studied. The results showed that RLD and RAI differed between genotypes; hybrid CRI46 had significantly higher （P 〈0.05） RLD and RAI values than conventlonal cultlvars, especially under low planting densities, less than 3.0 plants/m^2. The root area index （RAI） of hybrid CRI46 was 61% higher than of CRI44 and CRI45 at the flowering stage. The RLD and RAI were also significantly different （P = 0.000） between planting densities. The depth distribution of RAI showed that at Increasing planting densities RAI was Increasingly distributed in the soil layers below 50 cm. The RAI of hybrid CRI46 was for all planting densities, obviously higher than other cultivars during the flowering and boll stages. It was concluded that the hybrid had a strong advantage in root maintenance preventing premature senescence of roots. The root diameter of hybrid CRI46 had a genetically higher root diameter at planting densities lower than 6.0 plants/m^2. Good associations were found between yield and RAI In different stages. The optimum planting density ranged from 4.50 plants/m^2 to 6.75 plants/m^2 for conventional cultlvars and around 4.0-5.0 plants/m^2 for hybrids.

Effect of Long Protocol on the Outcome of In Vitro Fertilization-embryo Transfer in Patients with Endometriosis Compared with Prolonged Protocol

Objective To investigate the effect of different down-regulation protocol on the in vitro fertilization-embryo transfer （IVF-ET） outcomes in infertile patients with endometriosis （EMs）.Methods A retrospective case control study was performed. Totally 294 infertile patients with EMs were enrolled. And 109 patients （116 cycles） received prolongedprotocol as the control, 185 patients （193 cycles） received long protocol as case group, all followed by standard controlled ovarian hyperstimulation （COH）. Response to gonadotropins, the fertilization rate, the cleavage rate, the implantation rate, the clinical pregnancy rate and the miscarriage rate were measured and analyzed between the two groups. Results A trend toward better ovarian response was observed in long protocol group. Higher fertilization rate, lower total dose of rFSH, shorter duration of stimulation and more endometrial thickness on the day of hCG injection were observed in long protocol group compared with those of prolonged protocol group, and the difference was significant （P〈0.05）. In addition, the clinical pregnancy rate, the cleavage rate and the implantation rate also had an increase trend in long protocol group compared with those of prolonged protocol group, but without significant differences. Conclusion Long protocol regimen before IVF-ET in patients with EMs resulted in a trend toward better ovarian response and higher clinical pregnancy rates than prolonged protocol regimen did.