LiveZilla Live Chat Software

Institute for Capacity Development (ICD) is a non partisan independent Management training and development institution. ICD was founded on the need to continuously capacitate the skills of decision makers in Government, Parastatals, NGOs, the Private sector, CBOs and all other development based institutions.

The requisite skills and important for correct implementation of policy which will ultimately matter in the Social and Economic growth of Economies and the reduction of poverty and empowerment of people. All ICD programs take cognizance of the dynamic nature of different fields and hence the need to come up with contemporary solutions to problems using current methodologies. Information Technology Skills and the Application of Computers are mainstreamed in almost all ICD programs.

Top Panel
Saturday, 22 February 2020

Live support




LiveZilla Live Chat Software

LiveZilla Live Chat Software

Click here and chat to one of our Coordination Consultants

Contact Us Via Whatsapp and Viber

+27603048492; +264813866883
Whatsapp and Viber
 

Download Newsletter

ICD Newsletter

ICD Online Application

Multi-agent learning is primarily the intersection of two sub fields of artificial intelligence namely multi-agent systems and machine learning. Machine learning is being explored as one of the vital components to address the challenges in multi-agent systems. There are many application domains envisioned in which the machines learn to cooperate with each other and with human beings to achieve global objectives. The multi-agent learning is also vital to the learning process in non-cooperative domains such as finance and economics, where the traditional game-theoretic solutions are either unsuitable or infeasible. 
Multi-agent learning

Nonetheless, the multi-agent learning also poses a great deal of theoretical challenges in terms of understanding how the agents can learn and adapt in the presence of other agents that are adapting and learning simultaneously. This is a fertile area of research that seems ripe for progress such as decision theoretic and evolutionary learning that can be extended to more challenging multi-agent scenarios. 

Multi-agent Learning with Policy Prediction

The short courses, trainings and workshops on practice and theory in multi-agent learning are designed to be informal in style and broad in scope. The aim is to bring together researchers with a wide variety of perspectives who are look forward to take advantage of the open communications, common challenges, making comparison between the methodologies and sharing insights into recent results and future opportunities for progress in the field. The multi-agent learning offered by the Institute for Capacity Development covers a number of short talks by invited speakers as well as contributed talks and open discussion. 

Multi-agent learning is done by several agents and it becomes possible only because several agents are there. In true sense, if an agent looks forward to acquire skills to interact with other agents in its environment, the agent’s learning is called multi-agent learning, no matter whether or not other agents are involved in learning simultaneously.  Thus, it is possible to engage in multi-agent learning if only one agent is learning. The behavior is a multi-agent behavior, if the learned behavior supports additional multi-agent behaviors and in this more than one agent learns. The course covered in multi-agent learning at ICD staunchly believes that the learning would not be possible if the agents are isolated. 

The traditional machine learning involves only one agent trying to maximize some utility function without having knowledge or taking care of whether other agents are present in the environment or not. For instance, classification, function approximation and improvement in the performance of problem solving. 

The sub field of the multi-agent systems deals with the domains that have a number of agents and the mechanism for the interaction of independent behavior of agents. With the help of the multi-agent includes any situation in which an agent learns to interact with other agents effectively, regardless of whether the behavior of the agents is static.

The justification for the taking into account the situations in which only one agent learns to be multi-agent learning is that the learned behavior is the foundation for more complex and sophisticated interactive behaviors. In fact, though it seems that only one agent is involved in learning, it is only feasible if other agents are present and most importantly, it gives the agent an opportunity to participate in learning situations which are adverse and require high level of collaboration. In a bid to accomplish the multi-agent learning, layering of the learned behavior is essential and it involves the interaction with other agents.

Multi-agent Learning Techniques

The multi-agent learning program at ICD considers multi-agent learning as it involves an advisory and a cooperating agent which learn to interact with each other. This, in fact, is a negotiation technique. The situation only seems to be sensible it multiple agents are present. The multi agents attempt to model other agents. ICD emphasizes that multi-agent learning is a training scenario wherein a novice gets an opportunity to gain knowledge from a knowledgeable agent. 

The common thing between all the multi-agent learning systems taught at ICD is the interaction of the learning agent with other agents. Thus, it is emphasized that the learning process is feasible in the process of other agents and there are chances of interactions of higher level with these agents. 

 

Accreditation | Memberships | Partnership

ICDconsultancy Updated course list available on our website, please use the link below for more information and to be sure not to… https://t.co/AQQtkAowQA
ICDconsultancy Institute for Capacity Development (ICD) courses scheduled to start in February 2019 in South Africa. Book your se… https://t.co/KNxlAGB13m
ICDconsultancy Institute for Capacity Development (ICD) courses scheduled to start in January, February and March 2019 in South Af… https://t.co/3ld47xwfuA
ICDconsultancy Institute for Capacity Development (ICD) courses scheduled to start in December 2018 in South Africa. For more inf… https://t.co/flXsgj8MNy

International Institutions Trained

white_search.png

Search for 2020 Short Course


white_important.png

Upcoming 2020 Workshops



02nd – 13th March 2020
Training Needs Analysis and Staff Development
Pretoria - South Africa

02nd – 13th March 2020
Effective Purchasing, Tendering and Supplier Selection
Windhoek - Namibia

02nd – 13th March 2020
Leadership Skills for Supervisors - Communication, Coaching and Conflict
Pretoria - South Africa

02nd – 13th March 2020
Human Resources Management and Development
Pretoria - South Africa

02nd – 13th March 2020
Policy Formulation, Implementation and Evaluation
Pretoria - South Africa

02nd – 13th March 2020
Advanced Executive Office Administration and Secretarial Skills
Pretoria - South Africa

02nd – 13th March 2020
Effective Office Administration and Management
Harare - Zimbabwe

02nd – 13th March 2020
Results Based Monitoring and Evaluation of Development Projects
Harare - Zimbabwe

16th – 27th March 2020
Advanced Project Management
Kigali - Rwanda

16th – 27th March 2020
Public Sector Financial Management and control
Windhoek - Namibia

16th – 27th March 2020
Financial Management and Budgetary Control
Kigali - Rwanda

13th – 24th April 2020
Climate Change and Policy Design
Pretoria - South Africa

MORE COURSES.......