In this paper, we propose a new tractable framework for dealing with linear dynamical systems affected by uncertainty, applicable to multi-stage robust optimization and stochastic programming. We introduce a hierarchy of near-optimal polynomial disturbance-feedback control policies, and show how these can be computed by solving a single semidefinite programming problem. The approach yields a hi...

Running experiments on physical robotic systems typically requires significant resources (personnel, time, and money), restricting the number that can be run during training. Additionally, these systems typically operate over a range of external conditions and can be highly non-linear. We focus on the problem of learning a globally optimal policy to adapt controllers for such systems based on t...

Cache replacement policies are developed to help insure optimal use of limited resources. Varieties of such algorithms exist with relatively few that dynamically adapt to traffic patterns. Algorithms that are tunable typically utilize offline training mechanisms or trial-and-error to determine optimal characteristics. Utilizing multiple algorithms to establish an efficient replacement policy th...

Ensemble algorithms can improve the performance of a given learning algorithm through the combination of multiple base classifiers into an ensemble. In this paper we attempt to train and combine the base classifiers using an adaptive policy. This policy is learnt through a Q-learning inspired technique. Its effectiveness for an essentially supervised task is demonstrated by experimental results...

xiii

Effective fault-handling in emerging complex applications in large-scale MAS (Multi-agent Systems) requires the ability to dynamically adapt resource allocation and fault tolerance policies in response to changes in environment, user or application requirements, and available resources. This adaptation process incorporates an observation mechanism that transparently monitors the application’s b...

A new transaction model, called the Adaptive Transaction Model, is proposed for Real-Time Database Management Systems (RTDBMSs) applications. The Adaptive Transaction Model is an extended transaction model with a nested structure containing optional and required subtransactions. Adaptive Transactions (ATs) have time constraints to support real-time database applications. Optional substransactio...

− Dynamic load sharing policies take system state into account in making load distribution decisions. The state information can be maintained in one of two basic ways: distributed or centralized. Two principal types of policies that belong to the distributed scheme are the sender-initiated and receiverinitiated policies. In the centralized scheme, a central coordinator node is responsible for c...