Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Convergence of Probability Measures. By P. Billingsley. Chichester, Sussex, Wiley, 1968. xii, 253 p. 9 1/4“. 117s.

Convergence of Probability Measures

https://sci2s.ugr.es/sites/default/files/files/Teaching/GraduatesCourses/Metaheuristicas/Bibliography/VNS.pdf

Variable neighborhood search

The highly successful architecture and protocols of today's Internet may operate poorly in environments characterized by very long delay paths and frequent network partitions. These problems are exacerbated by end nodes with limited power or memory resources. Often deployed in mobile and extreme environments lacking continuous connectivity, many such networks have their own specialized protocols, and do not utilize IP. To achieve interoperability between them, we propose a network architecture and application interface structured around optionally-reliable asynchronous message forwarding, with limited expectations of end-to-end connectivity and node resources. The architecture operates as an overlay above the transport layers of the networks it interconnects, and provides key services such as in-network data storage and retransmission, interoperable naming, authenticated forwarding and a coarse-grained class of service.

/pdf/a-delay-tolerant-network-architecture-for-challenged-4nomv5ym6e.pdf

A delay-tolerant network architecture for challenged internets

An Introduction to MultiAgent Systems.

We consider an integrated sequencing and scheduling problem arising at filling lines in dairy industry. Even when a processing sequence is decided, still a scheduling problem has to be solved for the sequence. This incorporates typical side constraints as they occur also in other sequencing problems in practice. Previously, we proposed a framework for general sequencing and scheduling problems: A genetic algorithm is utilized for the sequencing, incorporating a problem specific algorithm for the fixed-sequence scheduling. In this paper, we investigate how this approach performs for filling lines. Based on insights into structural properties of the problem, we propose different scheduling algorithms. In cooperation with Sachsenmilch GmbH, the algorithm was implemented for their bottleneck filling line, and evaluated in an extensive computational study. For the real data from production, our algorithm computes almost optimal solutions. However, as a surprising result, our simple greedy algorithms outperform the more elaborate ones in many aspects, showing interesting directions for future research.

Sequencing and scheduling for filling lines in dairy production

Almost all comparability graphs are UPO

We consider the tasks (A) of minimizing a regular performance measure subject to resource constraints, and (B) of minimizing costs for resource requirements subject to a fixed completion time for arbitrary project networks with resource requirements. For these tasks we investigate those problems in which the optimum value is determined by the same partial order on the set of activities for all possible variations of the objective function (problems with a singular solution). It is shown that these problems constitute a large class of scheduling problems which can be recognized and solved in polynomial time. Furthermore, we give a complete characterization of all problems having a singular solution which is based on comparability graph recognition and forbidden substructures.

Scheduling Problems with a Singular Solution

Scheduling of partially ordered unit time jobs on m machines, aiming at minimal schedule length, is known to be one of the notorious combinatorial optimization problems, for which the complexity status is still unresolved. Availabe results give polynomial time algorithms for special classes of partial orders and for the two machine case. The present paper includes a general representation theorem on the existence of schedules of a given schedule length and a corresponding algorithmic method that allows a unifying approach to some of the solved special cases. It includes also a general algorithmic scheme that may work in the 3-machine case and may eventually lead to a polynomial time algorithm (or at least to membership in CO-NP) for the general m-machine case. Finally, it gives some insight into so-called m-critical posets, which should play an essential role in a proof on the correctness of the intended polynomial time algorithm.

M-Machine Unit Time Scheduling: A Report on Ongoing Research

Scheduling precedence constrained jobs on identical parallel machines is a well investigated problem with many applications. AND/OR-networks constitute a useful generalization of standard precedence constraints where certain jobs can be executed as soon as at least one of their direct predecessors is completed. For the problem of scheduling AND/OR-networks on parallel machines, we present a 2-approximation algorithm for the objective of minimizing the makespan. The main idea of the algorithm is to transform the AND/OR constraints into standard constraints. For the objective of minimizing the total weighted completion time on one machine, scheduling AND/OR-networks is as hard to approximate as Label Cover. We show that list scheduling with shortest processing time rule is an \(O(\sqrt{n})\)-approximation for unit weights on one machine and an n-approximation for arbitrary weights.

Rolf H. Möhring

Papers

Sequencing and scheduling for filling lines in dairy production

Almost all comparability graphs are UPO

Scheduling Problems with a Singular Solution

M-Machine Unit Time Scheduling: A Report on Ongoing Research

Scheduling AND/OR-networks on identical parallel machines