scholarly journals Human-level performance in 3D multiplayer games with population-based reinforcement learning

Science ◽  
2019 ◽  
Vol 364 (6443) ◽  
pp. 859-865 ◽  
Author(s):  
Max Jaderberg ◽  
Wojciech M. Czarnecki ◽  
Iain Dunning ◽  
Luke Marris ◽  
Guy Lever ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Video Game ◽  
Single Agent ◽  
Three Dimensional ◽  
Population Based ◽  
Great Success ◽  
Multiplayer Games ◽  
Multiagent Reinforcement Learning ◽  
The World ◽  
Level Performance

Reinforcement learning (RL) has shown great success in increasingly complex single-agent environments and two-player turn-based games. However, the real world contains multiple agents, each learning and acting independently to cooperate and compete with other agents. We used a tournament-style evaluation to demonstrate that an agent can achieve human-level performance in a three-dimensional multiplayer first-person video game, Quake III Arena in Capture the Flag mode, using only pixels and game points scored as input. We used a two-tier optimization process in which a population of independent RL agents are trained concurrently from thousands of parallel matches on randomly generated environments. Each agent learns its own internal reward signal and rich representation of the world. These results indicate the great potential of multiagent reinforcement learning for artificial intelligence research.

Scanning electron microscopy of freeze-fractured prescapular lymph node of caprine

1984 ◽  
Vol 42 ◽  
pp. 244-245
Author(s):  
O. Faroon ◽  
F. Al-Bagdadi ◽  
T. G. Snider ◽  
C. Titkemeyer
Keyword(s):  
Lymph Node ◽  
Lymph Nodes ◽  
Lymphatic System ◽  
Three Dimensional ◽  
Meat Products ◽  
The Body ◽  
Morphological Data ◽  
The World ◽  
Cellular Constituent ◽  
Scanning Electron

The lymphatic system is very important in the immunological activities of the body. Clinicians confirm the diagnosis of infectious diseases by palpating the involved cutaneous lymph node for changes in size, heat, and consistency. Clinical pathologists diagnose systemic diseases through biopsies of superficial lymph nodes. In many parts of the world the goat is considered as an important source of milk and meat products.The lymphatic system has been studied extensively. These studies lack precise information on the natural morphology of the lymph nodes and their vascular and cellular constituent. This is due to using improper technique for such studies. A few studies used the SEM, conducted by cutting the lymph node with a blade. The morphological data collected by this method are artificial and do not reflect the normal three dimensional surface of the examined area of the lymph node. SEM has been used to study the lymph vessels and lymph nodes of different animals. No information on the cutaneous lymph nodes of the goat has ever been collected using the scanning electron microscope.

scholarly journals A Population-Based Retrospective Study of Biliary Tract Cancers in Alberta, Canada

2021 ◽  
Vol 28 (1) ◽  
pp. 417-427
Author(s):  
Carissa Beaulieu ◽  
Arthur Lui ◽  
Dimas Yusuf ◽  
Zainab Abdelaziz ◽  
Brock Randolph ◽  
...  
Keyword(s):  
Biliary Tract ◽  
Demographic Data ◽  
Performance Status ◽  
Single Agent ◽  
Eastern Cooperative Oncology Group ◽  
Intrahepatic Bile Duct ◽  
Population Based ◽  
Phase Iii ◽  
First Line ◽  
Biliary Tract Cancers

Background: Biliary tract cancers (BTC) are uncommon malignancies and are underrepresented in the literature. Methods: We performed a retrospective population-based review of adult patients with biopsy-confirmed BTC in Alberta from 2000 to 2015. Demographic data, risk factors, symptoms, treatment, and staging data were collected and analyzed. Survival analyses were completed. Results: A total of 1604 patients were included in our study, of which 766 (47.8%) were male. The median age at diagnosis was 68 (range 19–99). There were 374 (23.3%) patients with resectable tumors at diagnosis versus 597 (37.2%) with unresectable tumors. Of the patients, 380 (21.5%) received chemotherapy (CT) and 81 (5.0%) underwent radiation therapy. There was a clear trend with worsening stage and performance status associated with shorter median overall survival (OS). Ampulla of Vater tumors had the best median OS (25.69 months), while intrahepatic bile duct cancers had the worst (5.78 months). First-line palliative CT regimens included gemcitabine+cisplatin (OS 14.98 months (mo), n = 212), single agent gemcitabine (OS 12.42 mo, n = 22), capecitabine (OS 8.12 mo, n = 8), and capecitabine+gemcitabine (OS 6.93 mo, n = 13). Patients with advanced or metastatic disease who received first-line gemcitabine+cisplatin had a median OS of 11.8 months (n = 119). Conclusion: BTCs have poor survival. Worse outcomes occur in higher stage and poorer Eastern Cooperative Oncology Group (ECOG) performance status patients across all tumor subtypes. Tumor resectability at diagnosis was associated with better OS. Our study supports the use of gemcitabine+cisplatin as a combination first-line palliative CT, as patients treated in Alberta have a comparable OS to that reported in the ABC-02 phase III study.

scholarly journals Multi-Agent Reinforcement Learning: A Review of Challenges and Applications

2021 ◽  
Vol 11 (11) ◽  
pp. 4948
Author(s):  
Lorenzo Canese ◽  
Gian Carlo Cardarilli ◽  
Luca Di Di Nunzio ◽  
Rocco Fazzolari ◽  
Daniele Giardino ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Mathematical Models ◽  
Learning Algorithms ◽  
Single Agent ◽  
Critical Issues ◽  
Multi Agent ◽  
Pros And Cons ◽  
Application Fields

In this review, we present an analysis of the most used multi-agent reinforcement learning algorithms. Starting with the single-agent reinforcement learning algorithms, we focus on the most critical issues that must be taken into account in their extension to multi-agent scenarios. The analyzed algorithms were grouped according to their features. We present a detailed taxonomy of the main multi-agent approaches proposed in the literature, focusing on their related mathematical models. For each algorithm, we describe the possible application fields, while pointing out its pros and cons. The described multi-agent algorithms are compared in terms of the most important characteristics for multi-agent reinforcement learning applications—namely, nonstationarity, scalability, and observability. We also describe the most common benchmark environments used to evaluate the performances of the considered methods.

scholarly journals Mortality beyond emergency threshold in a silent crisis– results from a population-based mortality survey in Ouaka prefecture, Central African Republic, 2020

2021 ◽  
Vol 15 (1) ◽  
Author(s):  
Eve Robinson ◽  
Lawrence Lee ◽  
Leslie F. Roberts ◽  
Aurelie Poelhekke ◽  
Xavier Charles ◽  
...  
Keyword(s):  
Central African Republic ◽  
Causes Of Death ◽  
Maternal Mortality Ratio ◽  
Mortality Rates ◽  
Population Based ◽  
Access To Healthcare ◽  
Mortality Data ◽  
Humanitarian Emergency ◽  
The World ◽  
Mortality Survey

Abstract Background The Central African Republic (CAR) suffers a protracted conflict and has the second lowest human development index in the world. Available mortality estimates vary and differ in methodology. We undertook a retrospective mortality study in the Ouaka prefecture to obtain reliable mortality data. Methods We conducted a population-based two-stage cluster survey from 9 March to 9 April, 2020 in Ouaka prefecture. We aimed to include 64 clusters of 12 households for a required sample size of 3636 persons. We assigned clusters to communes proportional to population size and then used systematic random sampling to identify cluster starting points from a dataset of buildings in each commune. In addition to the mortality survey questions, we included an open question on challenges faced by the household. Results We completed 50 clusters with 591 participating households including 4000 household members on the interview day. The median household size was 7 (interquartile range (IQR): 4—9). The median age was 12 (IQR: 5—27). The birth rate was 59.0/1000 population (95% confidence interval (95%-CI): 51.7—67.4). The crude and under-five mortality rates (CMR & U5MR) were 1.33 (95%-CI: 1.09—1.61) and 1.87 (95%-CI: 1.37–2.54) deaths/10,000 persons/day, respectively. The most common specified causes of death were malaria/fever (16.0%; 95%-CI: 11.0–22.7), violence (13.2%; 95%-CI: 6.3–25.5), diarrhoea/vomiting (10.6%; 95%-CI: 6.2–17.5), and respiratory infections (8.4%; 95%-CI: 4.6–14.8). The maternal mortality ratio (MMR) was 2525/100,000 live births (95%-CI: 825—5794). Challenges reported by households included health problems and access to healthcare, high number of deaths, lack of potable water, insufficient means of subsistence, food insecurity and violence. Conclusions The CMR, U5MR and MMR exceed previous estimates, and the CMR exceeds the humanitarian emergency threshold. Violence is a major threat to life, and to physical and mental wellbeing. Other causes of death speak to poor living conditions and poor access to healthcare and preventive measures, corroborated by the challenges reported by households. Many areas of CAR face similar challenges to Ouaka. If these results were generalisable across CAR, the country would suffer one of the highest mortality rates in the world, a reminder that the longstanding “silent crisis” continues.

scholarly journals Toxicity of Combinations of Kinase Pathway Inhibitors to Normal Human Cells in a Three-Dimensional Culture

2021 ◽  
pp. 247263032110088
Author(s):  
Pouria Rafsanjani Nejad ◽  
Pradip Shahi Thakuri ◽  
Sunil Singh ◽  
Astha Lamichhane ◽  
Jacob Heiss ◽  
...  
Keyword(s):  
Bone Marrow ◽  
Clinical Trials ◽  
Protein Kinase ◽  
Single Agent ◽  
Three Dimensional ◽  
Drug Combinations ◽  
Toxic Effects ◽  
Combination Drug ◽  
Normal Cells ◽  
Colon Cells

Resistance to single-agent chemotherapy and molecularly targeted drugs prevents sustained efficacy of treatments. To address this challenge, combination drug treatments have been used to improve outcomes for patients. Potential toxicity of combination treatments is a major concern, however, and has led to the failure of several clinical trials in different cancers. The use of cell-based models of normal tissues in preclinical studies enables testing and identifying toxic effects of drug combinations and facilitates an informed decision-making process for advancing the treatments to animal models and clinical trials. Recently, we established that combinations of molecular inhibitors of mitogen-activated protein kinase (MAPK) and phosphatidylinositol-3-kinase–protein kinase B (PI3K/Akt) pathways effectively and synergistically inhibit growth of BRAFmut and KRASmut colorectal tumor spheroids by blocking feedback signaling of downstream kinase pathways. These pathways are important for cell proliferation, however, and their simultaneous inhibition may cause toxicity to normal cells. We used a cellular spheroid model to study toxicities of drug combinations to human bone marrow and colon. Our results indicated that MAPK and PI3K/Akt inhibitors used simultaneously were only moderately toxic to bone marrow cells but significantly more toxic to colon cells. Our molecular analysis of proliferative cell activities and housekeeping proteins further corroborated these results. Overall, our approach to identify toxic effects of combinations of cancer drugs to normal cells in three-dimensional cultures will facilitate more informed treatment selections for subsequent animal studies.

scholarly journals On-Demand Channel Bonding in Heterogeneous WLANs: A Multi-Agent Deep Reinforcement Learning Approach

Sensors ◽  
2020 ◽  
Vol 20 (10) ◽  
pp. 2789 ◽  
Author(s):  
Hang Qi ◽  
Hao Huang ◽  
Zhiqun Hu ◽  
Xiangming Wen ◽  
Zhaoming Lu
Keyword(s):  
Reinforcement Learning ◽  
Transmission Rate ◽  
Single Agent ◽  
Time Of Day ◽  
Action Space ◽  
Traffic Load ◽  
Traffic Demand ◽  
Channel Bonding ◽  
On Demand ◽  
Multi Agent

In order to meet the ever-increasing traffic demand of Wireless Local Area Networks (WLANs), channel bonding is introduced in IEEE 802.11 standards. Although channel bonding effectively increases the transmission rate, the wider channel reduces the number of non-overlapping channels and is more susceptible to interference. Meanwhile, the traffic load differs from one access point (AP) to another and changes significantly depending on the time of day. Therefore, the primary channel and channel bonding bandwidth should be carefully selected to meet traffic demand and guarantee the performance gain. In this paper, we proposed an On-Demand Channel Bonding (O-DCB) algorithm based on Deep Reinforcement Learning (DRL) for heterogeneous WLANs to reduce transmission delay, where the APs have different channel bonding capabilities. In this problem, the state space is continuous and the action space is discrete. However, the size of action space increases exponentially with the number of APs by using single-agent DRL, which severely affects the learning rate. To accelerate learning, Multi-Agent Deep Deterministic Policy Gradient (MADDPG) is used to train O-DCB. Real traffic traces collected from a campus WLAN are used to train and test O-DCB. Simulation results reveal that the proposed algorithm has good convergence and lower delay than other algorithms.

Sign in / Sign up

Export Citation Format

Share Document