Luca Frigau

Light, temperature, dry after-ripening and salt stress effects on seed germination of Phleum sardoum (Hackel) Hackel

Phleum sardoum is an endemic psammophilous species of Sardinia, growing exclusively on coastal sandy dunes. The effect of glumes on seed germination, germination requirements at constant (5–25°C) and alternating (25/10°C) temperatures, both in the light (12/12 h) and in the dark were evaluated, as well as the effect of a dry after-ripening period (90 days at 25°C), the salt stress effect (0–600 mmol NaCl) and its recovery on seed germination. The presence of glumes reduced final germination percentages. For fresh naked seeds, high germination percentages were observed at 10°C. Dry after-ripening increased germination rate at low temperatures, but did not affect final germination percentages. NaCl determined a secondary salt-induced dormancy which recovery interrupted only partially. Our results highlighted that this species has its optimum of germination during autumn–winter when, under a Mediterranean climate, water availability is highest and soil salinity levels are minimal.

A data mining approach to forecast late arrivals in a transhipment container terminal

AbstractOne of the most important issues in Transhipment Container Terminal (TCT) management is to have fairly reliable and affordable predictions about vessel arrival. Terminal operators need to estimate the actual time of arrival in port in order to determine the daily demand for each work shift with greater accuracy. In this way, the resources required (human resources, equipment as well as spatial resources) can be allocated more efficiently. Despite contractual obligations to notify the Estimated Time of Arrival (ETA) 24 hours before arrival, ship operators often have to revise it due to unexpected events like weather conditions, delay in a previous port and so on. For planners the decision-making processes related to this topic can sometimes be so complex without the support of suitable methodological tools. Specific models should be adopted, in a daily planning scenario, to provide a useful support tool in TCTs. In this study, we discuss an exploratory analysis of the data affecting delays register...

Comparison of Cluster Analysis Approaches for Binary Data

Cluster methods allow to partition observations into homogeneous groups. Standard cluster analysis approaches consider the variables used to partition observations as continuous. In this work, we deal with the particular case all variables are binary. We focused on two specific methods that can handle binary data: the monothetic analysis and the model-based co-clustering. The aim is to compare the outputs performing these two methods on a common dataset, and figure out how they differ. The dataset on which the two methods are performed is a UNESCO dataset made up of 58 binary variables concerning the ability of UNESCO management to use Internet to promote world heritage sites.

Responsiveness and minimal important change of the NeckPix© in subjects with chronic neck pain undergoing rehabilitation

PurposeThe NeckPix© is a simple and rapid means of measuring the beliefs of subjects with chronic neck pain concerning pain-related fears of a specific set of activities of daily living. The original version showed satisfactory psychometric properties. This observational study is aimed at evaluating its responsiveness and minimal important changes (MICs) in subjects with chronic neck pain.MethodsAt the beginning, at the end of an 8-week rehabilitation programme as well as at the one-year follow-up, 153 subjects completed the NeckPix©. After the programme and at follow-up, subjects and physiotherapists also completed the global perceived effect (GPE) scale, which was divided to produce a dichotomous outcome. Responsiveness was calculated by distribution [effect size (ES); standardised response mean (SRM)] and anchor-based methods [receiver-operating characteristics (ROC) curves; correlations between change scores of the NeckPix© and GPEs]. ROC curves were also used to compute MICs.ResultsThe ES ranged from 0.95 to 1.26 and the SRM from 0.84 to 0.98 at post-treatment and follow-up based on subjects’ and physiotherapists’ perspective. The ROC analyses revealed AUCs of 0.89 and 0.97 at post-treatment and follow-up, respectively; MICs (sensitivity; specificity) were of 6 (0.82; 0.88) and 8 (0.80; 0.92) at post-treatment and of 8 (0.95; 0.90 based on subjects and 0.95; 0.92 based on physiotherapists perspective) at follow-up. The correlations between change scores of the NeckPix© and GPEs ranged from −0.69 to −0.82.ConclusionsThe NeckPix© was sensitive in detecting clinical changes in subjects with chronic neck pain undergoing rehabilitation. We recommend taking the MICs provided into account when assessing subjects’ improvement or planning studies in this clinical context.

Assessing the Reliability of a Multi-Class Classifier

Multi-class learning requires a classifier to discriminate among a large set of L classes in order to define a classification rule able to identify the correct class for new observations. The resulting classification rule could not always be robust, particularly when imbalanced classes are observed or the data size is not large. In this paper a new approach is presented aimed at evaluating the reliability of a classification rule. It uses a standard classifier but it evaluates the reliability of the obtained classification rule by re-training the classifier on resampled versions of the original data. User-defined misclassification costs are assigned to the obtained confusion matrices and then used as inputs in a Beta regression model which provides a cost-sensitive weighted classification index. The latter is used jointly with another index measuring dissimilarity in distribution between observed classes and predicted ones. Both indices are defined in [0, 1] so that their values can be graphically represented in a [0, 1]2 space. The visual inspection of the points for each classifier allows us to evaluate its reliability on the basis of the relationship between the values of both indices obtained on the original data and on resampled versions of it.