‘Horses for Courses’ in demand forecasting

Forecasting as a scientific discipline has progressed a lot in the last 40 years, with Nobel prizes being awarded for seminal work in the field, most notably to Engle, Granger and Kahneman. Despite these advances, even today we are unable to answer a very simple question, the one that is always the first tabled during discussions with practitioners: “what is the best method for my data?”. In essence, as there are horses for courses, there must also be forecasting methods that are more tailored to some types of data, and, therefore, enable practitioners to make informed method selection when facing new data. The current study attempts to shed light on this direction via identifying the main determinants of forecasting accuracy, through simulations and empirical investigations involving 14 popular forecasting methods (and combinations of them), seven time series features (seasonality, trend, cycle, randomness, number of observations, inter-demand interval and coefficient of variation) and one strategic decision (the forecasting horizon). Our main findings dictate that forecasting accuracy is influenced as follows: (a) for fast-moving data, cycle and randomness have the biggest (negative) effect and the longer the forecasting horizon, the more accuracy decreases; (b) for intermittent data, inter-demand interval has bigger (negative) impact than the coefficient of variation; and (c) for all types of data, increasing the length of a series has a small positive effect.     

A mushy region in concrete corrosion

In this work a mathematical model already known for the corrosion of sewer pipes is further considered, enriched and an approximate analytical solution is given based on a quasi-steady approximation. Furthermore an extension of this model is constructed allowing the formation of a mushy region, i.e. a region in which the material is only partly corroded throughout the process. This more general model is derived via an averaging process by microscopic considerations and has the form of a macroscopic phase field model. The derived model in its various forms is solved numerically with a finite element method and the results, predicting corrosion within a reasonable range, are presented.     

Modified YSZ/Ni structures with improved stability

The modification of the YSZ/Ni interface by the addition of lower surface energy oxides such as TiO₂, Cr₂O₃ and Mn₃O₄ on the YSZ-surface or alloying Ni-metal with Cr or Pd influences positively the interfacial bonding between ceramic and metallic phases. Wetting experiments in the modified-YSZ/liquid Ni and YSZ/liquid Ni-based alloys systems at 1773K in Ar+4vol%H₂ atmosphere gave contact angle values lower than that of the unmodified YSZ/liquid-Ni system. The present wettability results together with literature data were used for the determination of the surface energies for the modified YSZ-ceramics and the liquid Ni-based alloys as well as the interfacial energies of the ceramic-liquid metal systems examined. Lower interfacial energy values were achieved in the case of the modified YSZ-ceramics. The modification of the YSZ-surface was found to contribute positively also to the interfacial bonding of the ceramic-solid Ni interface, decreasing in this way the agglomeration rate of the metallic phase, when this is deposited as a thin layer onto the ceramic surface. Paper presented at the 3rd Euroconference on Solid State Ionics, Teulada, Sardinia, Italy, Sept. 16, 1996     

Direct versus sequential double ionization of Mg with extreme-ultraviolet radiation

We present an ab initio calculation of multiphoton double ionization in atomic magnesium, the first of its type in an atom with a complex core, showing that direct correlated double electron ejection can be studied through analysis of the photoelectron energy spectrum, using currently or upcoming extreme-ultraviolet radiation sources.     

Understanding the Effects of Binders in Gas Sorption and Acidity of Aluminium Fumarate Extrudates

Understanding the impact of shaping processes on solid adsorbents is critical for the implementation of MOFs in industrial separation processes or as catalytic materials. Production of MOF-containing shaped particles is typically associated with loss of porosity and modification of acid sites, two phenomena that affect their performance. Herein, we report a detailed study on how extrusion affects the crystallinity, porosity, and acidity of the aluminium fumarate MOF with clays or SiO₂ gel binders. Thorough characterization showed that the clay binders confer the extrudates a good mechanical robustness at the expense of porosity, while silica gel shows an opposite trend. The CO₂ selectivity towards CH₄, of interest for natural gas separation processes, is maintained upon the extrusion process. Moreover, probe FTIR spectroscopy revealed no major changes in the types of acid sites. This study highlights that these abundant and inexpensive clay materials may be used for scaling MOFs as active adsorbents.     

Honey Phenolic Compound Profiling and Authenticity Assessment Using HRMS Targeted and Untargeted Metabolomics

Honey consumption is attributed to potentially advantageous effects on human health due to its antioxidant capacity as well as anti-inflammatory and antimicrobial activity, which are mainly related to phenolic compound content. Phenolic compounds are secondary metabolites of plants, and their content in honey is primarily affected by the botanical and geographical origin. In this study, a high-resolution mass spectrometry (HRMS) method was applied to determine the phenolic profile of various honey matrices and investigate authenticity markers. A fruitful sample set was collected, including honey from 10 different botanical sources (n = 51) originating from Greece and Poland. Generic liquid–liquid extraction using ethyl acetate as the extractant was used to apply targeted and non-targeted workflows simultaneously. The method was fully validated according to the Eurachem guidelines, and it demonstrated high accuracy, precision, and sensitivity resulting in the detection of 11 target analytes in the samples. Suspect screening identified 16 bioactive compounds in at least one sample, with abscisic acid isomers being the most abundant in arbutus honey. Importantly, 10 markers related to honey geographical origin were revealed through non-targeted screening and the application of advanced chemometric tools. In conclusion, authenticity markers and discrimination patterns were emerged using targeted and non-targeted workflows, indicating the impact of this study on food authenticity and metabolomic fields.     

Stochastic and Self-Organisation Patterns in a 17-Year PM10 Time Series in Athens,Greece

This paper utilises statistical and entropy methods for the investigation of a 17-year PM₁₀ time series recorded from five stations in Athens, Greece, in order to delineate existing stochastic and self-organisation trends. Stochastic patterns are analysed via lumping and sliding, in windows of various lengths. Decreasing trends are found between Windows 1 and 3500–4000, for all stations. Self-organisation is studied through Boltzmann and Tsallis entropy via sliding and symbolic dynamics in selected parts. Several values are below −2 (Boltzmann entropy) and 1.18 (Tsallis entropy) over the Boltzmann constant. A published method is utilised to locate areas for which the PM₁₀ system is out of stochastic behaviour and, simultaneously, exhibits critical self-organised tendencies. Sixty-six two-month windows are found for various dates. From these, nine are common to at least three different stations. Combining previous publications, two areas are non-stochastic and exhibit, simultaneously, fractal, long-memory and self-organisation patterns through a combination of 15 different fractal and SOC analysis techniques. In these areas, block-entropy (range 0.650–2.924) is significantly lower compared to the remaining areas of non-stochastic but self-organisation trends. It is the first time to utilise entropy analysis for PM₁₀ series and, importantly, in combination with results from previously published fractal methods.Data Set License: license under which the dataset is made available (CC0, CC-BY, CC-BY-SA, CC-BY-NC, etc.)     

Dual-cardiac marker capillary waveguide fluoroimmunosensor based on tyramide signal amplification

The early diagnosis of acute myocardial infarction requires the determination of several markers in serum shortly after its incidence. The markers most widely employed are the isoenzyme MB of creatine kinase (CK-MB) and the cardiac troponin I (cTnI). In the present work, a capillary waveguide fluoroimmunosensor for fast and highly sensitive simultaneous determination of these markers in serum samples is demonstrated. The dual-analyte immunosensor was realized using glass capillaries internally modified with an ultrathin poly(dimethylsiloxane) film by creating discrete bands of analyte-specific antibodies. The capillary was then filled with a mixture of sample and biotinylated detection antibodies followed by reaction with streptavidin–horseradish peroxidase and incubation with a fluorescently labeled tyramide derivative to accumulate fluorescent labels onto immunoreaction bands. Upon scanning the capillary with a laser beam, part of the emitted fluorescence is trapped and waveguided through the capillary wall to a photomultiplier placed on one of its ends. The employment of tyramide signal amplification provided detection limits of 0.2 and 0.5 ng/mL for cTnI and CK-MB, respectively, in a total assay time of 30 min compared to 0.8 and 0.6 ng/mL obtained for the corresponding assays when the conventional fluorescent label R-phycoerythrin was used in a 65-min assay. In addition, the proposed immunosensor provided accurate and repeatable measurements (intra-assay and interassay coefficients of variation lower than 10%), and the values determined in serum samples were in good agreement with those obtained with commercially available enzyme immunoassays. Thus, the proposed capillary waveguide fluoroimmunosensor has all the required characteristics for fast and reliable diagnosis of acute myocardial infarction.      

Influence of the helium autoionization structures on the single/double ionization signal

Calculations of intense field (around 10¹⁶ W/cm²) single- and double-ionization processes in helium at XUV wavelengths are presented. The laser wavelength is chosen near the | 2s2p ¹ P autoionization structure and the dynamics are explored. Single and double ionization yields, as well as the photoelectron energy spectrum for photon energies around the autoionization structure are calculated. In the case of a pulse of few femtoseconds duration, no significant enhancement of the double ionization yield has been found in tuning the photon frequency around the peak of the resonance. It is also shown that in the case of a long pulse (and hence narrow compared with the relevant autoionization width), the branching ratio of double to single ionization yield can be relatively enhanced by tuning to the absorption minimum of the resonance. Received 19 February 2002 / Received in final form 2 May 2002 Published online 19 July 2002     

Excitation and ionization of positronium by 50–100 fs laser pulses

The interaction between Ps and a strong laser pulse of short duration is studied. Substantial population is deposited in excited states under few photon excitation. The possible usefulness as a source of excited Ps to spectroscopy and the formation of antihydrogen is discussed. Finally, photoelectron energy spectra are calculated. This revised version was published online in August 2006 with corrections to the Cover Date.