Feature selection methods for root-cause analysis among top-level product attributes

Manufacturing companies not only strive to deliver flawless products but also monitor product failures in the field to identify potential quality issues. When product failures occur, quality engineers must identify the root cause to improve any affected product and process. This root-cause analysis can be supported by feature selection methods that identify relevant product attributes, such as manufacturing dates with an increased number of product failures. In this paper, we present different methods for feature selection and evaluate their ability to identify relevant product attributes in a root-cause analysis. First, we compile a list of feature selection methods. Then, we summarize the properties of product attributes in warranty case data and discuss these properties regarding the challenges they pose for machine learning algorithms. Next, we simulate datasets of warranty cases, which emulate these product properties. Finally, we compare the feature selection methods based on these simulated datasets. In the end, the univariate filter information gain is determined to be a suitable method for a wide range of applications. The comparison based on simulated data provides a more general result than other publications, which only focus on a single use case. Due to the generic nature of the simulated datasets, the results can be applied to various root-cause analysis processes in different quality management applications and provide a guideline for readers who wish to explore machine learning methods for their analysis of quality data.     

Vergleichende Untersuchungen der trypsinspezifischen reaktiven Zentren verschiedener Phaseolus coccineus- und Phaseolus vulgaris-Inhibitoren

Zusammenfassung Drei Trypsin-Chymotrypsin-Inhibitoren aus den Samen von Feuerbohnen (Phaseolus coccineus L.) PCI 3, 4₂ und 5 und drei Inhibitoren aus den Samen von Buschbohnen (Phaseolus vulgaris var. nanus) PVI 3, 4 und 5 enthalten im reaktiven Zentrum gegen Trypsin einen Lysin-Rest. Ein Inhibitor ausPhaseolus coccineus, PCI 2, enthält dort einen Arginin-Rest. Alle sieben untersuchtenPhaseolus-Inhibitoren sinddouble headed.

---

Comparative studies on the reactive sites against trypsin of some inhibitors from Phaseolus coccineus and Phaseolus vulgaris

Summary Three of the trypsin chymotrypsin inhibitors from the seeds of runner beans (Phaseolus coccineus L.), PCI 3, 4₂, and 5, and three of the inhibitors from the seeds of french beans (Phaseolus vulgaris var. nanus), PVI 3, 4, and 5, contain a lysine residue in the reactive site against trypsin. One of the inhibitors fromPhaseolus coccineus, PCI 2, contains an arginine residue there. All sevenPhaseolus inhibitors investigated are double headed.

---

---

Die Arbeiten wurden durch die AIF über den Forschungskreis der Ernährungsindustrie gefördert     

Automated Karl Fischer titration for liquid samples – Water determination in edible oils

In the field of food analysis, rapid measurements and results are generally of high importance. The Karl Fischer titration (KFT), a chemical method for determining water content, and its automated performance enhance rapidity by giving the opportunity to deal with more samples in less time; automated sequences can include different determination methods and sample treatments. Moreover, automation can improve reproducibility and precision.     

Water determination – Scientific and economic dimensions

Water content is for a number of reasons one of the most important properties of foodstuffs. The determination of water content is therefore a very important analysis. This is not only the case for scientific or technical reasons. The water fraction in food is sometimes regarded as a component without commercial value. The price is based on dry matter, which makes water content determination an analysis with economic consequences. Different existing methods often yield different results. This entrains the problem that vendor and buyer, for economic reasons, prefer different methods to favour their role in trade. The situation would be easier if the method to be applied would be the same and would be agreed upon by both parties. International standards and reference methods are established by bodies that consist not only of neutral scientists but also of representatives of industry. These will usually see the interest of their respective company and will rather try to introduce methods which “improve” their profit rather than accepting methods scientifically correct but with lower profit. Milk powder trade is given as an example to illustrate such a situation.     

A system for simulation, monitoring and diagnosis of controlled continuous processes with slow dynamics

This paper presents a modelling approach specifically designed to be used in a monitoring and diagnosis system based on the simulation of relevant process and control system variables. The model is based on semantically extended data flow diagrams with each computational process representing the incremental behaviour of a component of the plant to be modelled. Variable as well as delay time behaviour is associated with each component model to support an event-driven simulation which, in turn, reduces computation time. This plant model of independent but interacting computational processes is easily extendable, modifiable and adjustable according to changes in the plant.

The modelling approach is applied to a coal power plant process and is implemented in G2, an object-oriented, real-time expert system shell with built-in simulation and online external connection capabilities. Generic rules guiding the event-driven simulation and monitoring allow for modular adaptation of the model as the plant's structure and functionality change without requiring changes in the rule set.