Backward elimination model construction for regression and classification using leave-one-out criteria

A fundamental principle in practical nonlinear data modeling is the parsimonious principle of constructing the minimal model that explains the training data well. Leave-one-out (LOO) cross validation is often used to estimate generalization errors by choosing amongst different network architectures (M. Stone, "Cross validatory choice and assessment of statistical predictions", J. R. Stast. Soc., Ser. B, 36, pp. 117-147, 1974). Based upon the minimization of LOO criteria of either the mean squares of LOO errors or the LOO misclassification rate respectively, we present two backward elimination algorithms as model post-processing procedures for regression and classification problems. The proposed backward elimination procedures exploit an orthogonalization procedure to enable the orthogonality between the subspace as spanned by the pruned model and the deleted regressor. Subsequently, it is shown that the LOO criteria used in both algorithms can be calculated via some analytic recursive formula, as derived in this contribution, without actually splitting the estimation data set so as to reduce computational expense. Compared to most other model construction methods, the proposed algorithms are advantageous in several aspects; (i) There are no tuning parameters to be optimized through an extra validation data set; (ii) The procedure is fully automatic without an additional stopping criteria; and (iii) The model structure selection is directly based on model generalization performance. The illustrative examples on regression and classification are used to demonstrate that the proposed algorithms are viable post-processing methods to prune a model to gain extra sparsity and improved generalization.

A radial basis function network classifier to maximise leave-one-out mutual information

tWe develop an orthogonal forward selection (OFS) approach to construct radial basis function (RBF)network classifiers for two-class problems. Our approach integrates several concepts in probabilisticmodelling, including cross validation, mutual information and Bayesian hyperparameter fitting. At eachstage of the OFS procedure, one model term is selected by maximising the leave-one-out mutual infor-mation (LOOMI) between the classifier’s predicted class labels and the true class labels. We derive theformula of LOOMI within the OFS framework so that the LOOMI can be evaluated efficiently for modelterm selection. Furthermore, a Bayesian procedure of hyperparameter fitting is also integrated into theeach stage of the OFS to infer the l2-norm based local regularisation parameter from the data. Since eachforward stage is effectively fitting of a one-variable model, this task is very fast. The classifier construc-tion procedure is automatically terminated without the need of using additional stopping criterion toyield very sparse RBF classifiers with excellent classification generalisation performance, which is par-ticular useful for the noisy data sets with highly overlapping class distribution. A number of benchmarkexamples are employed to demonstrate the effectiveness of our proposed approach.

Hälsofrämjande arbete på SSAB i Borlänge : Hälsofrämjande processer i arbetslivet och ledarskapets betydelse för medarbetarens hälsa

Problem: Arbetsmiljön är viktig för människors välbefinnande. Hälsofrämjande faktorer antas inte bara styrka individens fysiska och psykiska hälsa, utan även företagets konkurrenskraft och lönsamhet. I uppsatsen undersöker vi hur företaget SSAB i Borlänge arbetar med hälsa genom att tillämpa ett hälsofrämjande perspektiv och utgå från teorin om Känslan av sammanhang (KASAM). Ledarskapsvärderingar har betydelse för medarbetares hälsa (Hanson, 2004) och vi undersöker vilken uppfattning ett antal chefer har om sina roller i det hälsofrämjande arbetet och deras syn på medarbetarundersökningen HälsoSAM som företagshälsovården på SSAB bedriver. Arbetsbelastningen på medarbetare och chefer ökar till följd av sparkrav, samtidigt som resurserna minskar (Gatu, 2003). Följden blir en större risk för ohälsa och ett sätt att minska sjukfrånvaron är att genomföra hälsofrämjande insatser (Prevent, 2001). Syfte: Syftet med studien är att utifrån ett hälsofrämjande perspektiv förklara vad chefer har för möjligheter att skapa förutsättningar för att främja medarbetarnas hälsa på SSAB i Borlänge. Metod: Den metod som ligger till grund för uppsatsen baseras på ett kvalitativt angreppssätt där semistrukturerade intervjuer samt litteraturstudier genomförts för att samla information. Analys: Ledarens beteende påverkar medarbetarna. Hög arbetsbelastning på SSABs chefer leder till lägre närvaro bland medarbetarna, vilket i sin tur ger sämre förutsättningar för att främja hälsan. HälsoSAM kartlägger medarbetarnas hälsoläge, både välbefinnande och arbetskapacitet. Resultaten är vägledande i det hälsofrämjande arbetet men överbelastning, tidsbrist och kunskapsbrist hos cheferna bidrar till att uppföljningen inte blir systematisk. För ett väl fungerande hälsoarbete krävs systematik. Slutsats: Hälsofrämjande processer i arbetslivet skapas genom balans mellan krav och resurser. På SSAB i Borlänge hindras hälsoarbetet av tidsbrist och kunskapsbrist hos cheferna. Kostnadsbesparingar påverkar hälsoarbetet negativt och systematiken i uppföljningsarbetet blir lidande. Ansvaret för hälsoarbetet läggs på företagshälsovården, men utan stöd från chefer blir inte hälsoarbetet en naturlig del i verksamheten. Det är viktigt att i rådande situation se vad främjande av hälsa kan ge tillbaka till företaget i både ekonomiska och kvalitativa termer. Insikt om detta ökar chefernas incitament att prioritera hälsofrämjande processer och organisera arbetet så att människor har förutsättningar att hantera, kontrollera och klara av sina uppgifter.

Do you really want to ask an underwriter how much money you should leave on the table?

This paper analyses whether the owners of companies seeking to list will leave less money on the table if underwriters are employed to price and market the issue. Our findings indicate that limited liability and Industrial company initial public offerings (IPOs) that have used underwriters have left
more money on the table than those not employing underwriters. Not only is there a direct cost in employing an underwriter but this study suggests there might also be an indirect cost. We also find that a positive forecast earnings per share yield may be useful in reducing the amount of money left on the table.

Promise more, leave less

Following Kim and Ritter (1999) who find that earnings forecasts provide more accurate valuations of IPOs, this paper analyses whether the owners of companies seeking to list will leave less money on the table if positive dividend per share (DPS) yield forecasts are made in the prospectus. Our findings indicate that DPS yield forecasts by directors of Industrial company IPOs have been an important ingredient in the amount of money left on the table. A similar result is found for Limited Liability IPOs and those that do not offer options to subscribers to buy more shares. The offer of an operational dividend reinvestment plan in the prospectus does not appear significant in reducing the amount of money left on the table.

Portable sick leave in the Victorian building industry : managing cumulative employee benefits in the absence of employment security

The trend away from full-time permanent employment raises questions about the relevance of traditional approaches to managing and compensating employees. Employment in the Australian building industry is characterised by short-term, project-based employment. Employers and unions in the industry have adopted alternative compensation models to accommodate the short-term nature of employment, most notably through portable benefit schemes. In 1997, the Victorian building industry extended the range of portable benefits to include sick leave. Empirical evidence suggests a relationship between employee absence behaviour and accrual entitlement models. Research reported here supports this link, and suggests that both employers and employees can benefit from an alternative, portable, approach to accrued entitlements. Employers can benefit because employees may be less likely to take an instrumental approach to their entitlements. Employees benefit because they are able to accrue entitlements for the period they remain in the building industry, irrespective of the extent to which they change jobs.

Reactions to increased workload: effects on professional efficacy of nurses

Research findings support the idea that workload is a significant stressor associated with a variety of deleterious psychological reactions, including burnout, in several different samples of workers. A theoretical model is put forth in the present study in which workload is seen as contributing to distress and depression. Increasingly, organisations are experiencing changes as a result of extensive downsizing, restructuring, and merging. As a result of fiscal restraint, hospitals have been forced to merge, close, downsize, and restructure. Workloads have increased among hospital staff, particularly nurses. This study applies a theoretical model to the understanding of the impact of workload on nurses employed in hospitals experiencing downsizing, particularly on their distress, burnout, and depression. Respondents were 488 nurses who were employed in hospitals that were undergoing restructuring and in which units had already been closed as a result of restructuring. Results of structural equation modeling showed that the data partially fit the model and that workload contributed substantially to levels of depression through distress reactions. Further results showed that cynicism, anger, and emotional exhaustion significantly operationalised distress reactions. This study is unique theoretically in linking anger, cynicism, and emotional exhaustion in a single model that predicts distress levels from workload. The findings that anger, cynicism, and emotional exhaustion operationalised distress indicate the importance of studying patterns of negative reactions and their consequences for depression. Implications of the results are discussed for interventions that can be taken by organisations in order to reduce workloads.

Workload capacity measures for estimating allied health staffing requirements

Workforce planning methodologies for the allied health professions are acknowledged as rudimentary despite the increasing importance of these professions to health care across the spectrum of health services settings. The objectives of this study were to (i) identify workload capacity measures and methods for profiling allied health workforce requirements from a systematic review of the international literature; (ii) explore the use of these methods in planning workforce requirements; (iii) identify barriers to applying such methods; and (iv) recommend further action. Future approaches to workforce planning were explored through a systematic review of the literature, interviews with key stakeholders and focus group discussions with representatives from the different professional bodies and health agencies in Victoria. Results identified a range of methods used to calculate workload requirements or capacity. In order of increasing data demands and costliness to implement, workload capacity methods can be broadly classified into four groups: ratio-based, procedure-based, categories of care-based and diagnostic or casemix-based. Despite inherent limitations, the procedure-based measurement approach appears to be most widely accepted. Barriers to more rigorous workforce planning methods are discussed and future directions explored through an examination of the potential of casemix and mixed-method approaches.

Determining clinical pharmacy workload by patient disease classification in medical and surgical patients

Aim: To determine the time needed to provide clinical pharmacy services to individual patient episodes for medical and surgical patients and the effect of patient presentation and complexity on the clinical pharmacy workload. Method: During a 5-month period in 2006 at two general hospitals, pharmacists recorded a defined range of activities that they provided for patients, including the actual times required for these tasks. A customised database linked to the two hospitals' patient administration systems stored the data according to the specific patient episode number. The influence of patient presentation and complexity on the clinical pharmacy activities provided was also examined. Results: The average time required by pharmacists to undertake a medication history interview and medication reconciliation was 9.6 (SD 4.9) minutes. Interventions required 5.7 (SD 4.6) minutes, clinical review of the medical record 5.5 (SD 4.0) minutes and medication order review 3.5 (SD 2.0) minutes. For all of these activities, the time required for medical patients was greater than for surgical patients and greater for 'complicated' patients. The average time required to perform all clinical pharmacy activities for 1071 completed patient episodes was 14.4 (SD 10.9) minutes and was greater for medical and 'complicated' patients. Conclusion: The time needed to provide clinical pharmacy services was affected by whether the patients were medical or surgical. The existence of comorbidities or complications affected these times. The times required to perform clinical pharmacy activities may not be consistent with recently proposed staff ratios for the provision of a basic clinical pharmacy service.

The impact of workload on wellbeing, mental health and depression : a longitudinal study of work perception

This paper explores the relationships between characteristics of the job (workload, control and support) and organizational justice (distributive, procedural, interpersonal and informational) at Time 1, onto three indicators of psychological health at Time 2 (psychological wellbeing, distress and depression). The sample consisted of sworn members of a state-based police force (n=143). Hierarchical regression analyses indicated that workload was associated with psychological wellbeing, distress and depression at the one-year follow-up. Specifically, high workload at Time 1 was associated with psychological distress and depression at Time 2, and low workload was associated with psychological wellbeing at Time 2. Further, there was a significant relationship between perceived informational justice at Time 1 and psychological wellbeing at Time 2. No significant interaction effects were demonstrated for the job characteristics or organizational justice onto psychological health status. That is, longitudinally, workload directly influences both positive and negative mental health, and informational justice is related to psychological wellbeing. The implications for the demand-control-support model are discussed. The injustice-as-stressor argument was generally not supported.