The exact sample complexity of PAC-learning problems with unit VC dimension

The Vapnik-Chervonenkis (VC) dimension is a combinatorial measure of a certain class of machine learning problems, which may be used to obtain upper and lower bounds on the number of training examples needed to learn to prescribed levels of accuracy. Most of the known bounds apply to the Probably Approximately Correct (PAC) framework, which is the framework within which we work in this paper. For a learning problem with some known VC dimension, much is known about the order of growth of the sample-size requirement of the problem, as a function of the PAC parameters. The exact value of sample-size requirement is however less well-known, and depends heavily on the particular learning algorithm being used. This is a major obstacle to the practical application of the VC dimension. Hence it is important to know exactly how the sample-size requirement depends on VC dimension, and with that in mind, we describe a general algorithm for learning problems having VC dimension 1. Its sample-size requirement is minimal (as a function of the PAC parameters), and turns out to be the same for all non-trivial learning problems having VC dimension 1. While the method used cannot be naively generalised to higher VC dimension, it suggests that optimal algorithm-dependent bounds may improve substantially on current upper bounds.

Post-observation feedback as an instigator of teacher learning and change

This paper investigates an on-going study into the nature of the discourse teacher educators (TEds) and student-teachers (Ts) use during postobservation feedback, and explores the feedback discourse in terms of the impact it may have on teacher learning and change. Two types of feedback, confirmatory and corrective, are briefly discussed, and the former is then made the focus of the paper in its role as a potential instigator of teacher learning and change.

The views of selected NGOs on corporate social disclosures in Bangladesh

Although prior studies looked at corporate social disclosures (CSD hereafter) mainly from the managerial perspective there are very few studies which examined CSD from a non-managerial stakeholder perspective. This paper contributes to that limited CSD literature. It does so from a developing country perspective. The main aim of this paper is to examine the views of selected NGOs on current CSD practices in Bangladesh using Gramscian hegemonic analysis. For this purpose, semi-structured interviews were carried out in the selected social and environmental NGOs of both overseas and Bangladesh origin. The results suggest that NGOs viewed the current CSD practice as far from satisfactory. They also argued that it is mainly aimed at maintaining corporate interests of image building. The study suggests that it is not corporations to be blamed alone for production of CSD in the interests of business, it is the capitalist society that consents to such reproduction of CSD.

The views of corporate managers on the current state of, and future prospects for, social reporting in Bangladesh: an engagement based study

Purpose – This paper seeks to respond to recent calls for more engagement-based studies of corporate social reporting (CSR) practice by examining the views of corporate managers on the current state of, and future prospects for, social reporting in Bangladesh. Design/methodology/approach – The paper uses a series of interviews with senior managers from 23 Bangladeshi companies representing the multinational, domestic private and public sectors. Findings – Key findings are that the main motivation behind current reporting practice lies in a desire on the part of corporate management to manage powerful stakeholder groups, whilst perceived pressure from external forces, notably parent companies' instructions and demands from international buyers, is driving the process forward. In the latter context it appears that adoption of international social accounting standards and codes is likely to become more prevalent in the future. Reservations are expressed as to whether such a passive compliance strategy is likely to achieve much in the way of real changes in corporate behaviour, particularly when Western developed standards and codes are imposed without consideration of local cultural, economic and social factors. Indeed, such imposition could be regarded as little more than an example of the erection of non-tariff trade barriers rather than representing any meaningful move towards empowering indigenous stakeholder groups. Originality/value – The paper contributes to the literature on CSR in developing countries where there is a distinct lack of engagement-based published studies.

Recognition times of different views of 56 depth-rotated objects: a note concerning Verfaillie and Boutsen (1995)

The present study examines the effect of the goodness of view on the minimal exposure time required to recognize depth-rotated objects. In a previous study, Verfaillie and Boutsen (1995) derived scales of goodness of view, using a new corpus of images of depth-rotated objects. In the present experiment, a subset of this corpus (five views of 56 objects) is used to determine the recognition exposure time for each view, by increasing exposure time across successive presentations until the object is recognized. The results indicate that, for two thirds of the objects, good views are recognized more frequently and have lower recognition exposure times than bad views.