Problems solving

Problem-solving is the most fundamental of scientific tasks. At its heart is the search for an understanding of the behaviour of the physical world, which, in its most comprehensive form, brings together both descriptive and quantitative elements. It is the search for solutions to scientific problems which is the central theme of this volume. In the chapters that follow, chemists discuss recent progress in the application of some of the most promising algorithms now available to solve scientific problems. 

Within the rich set of scientific problems are many that are difficult to solve. It is clearly unrealistic to expect that any single method can invariably be successful, no matter what the type of problem to which it is applied, so scientists must be pragmatic, taking advantage of whatever algorithmic tools are available and seem likely to be productive. 

It is a characteristic of most iterative methods that their performance is strongly correlated with the values chosen for adjustable parameters. Thus, the step size in a hill-climb must be selected with care (and perhaps adjusted as the calculation proceeds) to ensure that movement across the search landscape is neither so languid that the task of locating the maximum is unreasonably drawn out, nor so volatile and unpredictable that the search is unable to settle on a maximum. Mathematical recipes sometimes exist which specify how the values of adjustable parameters in iterative algorithms should be chosen, or how they can be optimised as the calculation proceeds. 

Modeling is an exercise in problem solving and there are steps that can be taken to maximize the probability of solving the problem. The following outlines the steps for effective problem solving  

Analyze the problem. What data are available to solve the problem Given the data available can the problem be solved Has proper attention to study design and data collection been done to achieve the objective Question whether a model is even necessary. Perhaps a noncompartmental analysis of the data will suffice instead. If the goal is to model multiple-dose data from single dose data, then something simple like the superposition principle may be useful. 

Identify alternative solutions. Review past solutions for current problems. Perhaps something that you are trying to do has already been done and reported in the literature for a different drug. Sometimes it is possible to break a complex problem into a series of simpler problems. Sometimes you will need to be creative though and possibly need to brainstorm with others. 

Evaluate the possible solutions. Define criteria for choosing a solution. Is time more important than cost What is the most cost- and time-effective alternative to answer the question Understand who will use the results and what will be the best way to communicate those results. Perhaps modeling is not the optimal solution. 

Decide on a solution, keeping in mind that rarely will any single solution be perfect. Identify limitations of all proposed solutions. 

In the ideal world moulding faults do not exist. Unhappily companies continue to live with problems, either because the problem is thought to be impossible to solve or it has not been recognised as a production cost that can be saved. If company rejects are zero, then skip this section  

The first step in improving any process is to ensure that all product failures are counted and classified according to the reason for rejection. This process needs to be done on a regular shift-by-shift basis. 

If you assume that the numbers refer to the year of minting, the question seems to be nonsense as a dime is worth 10 cents no matter what the year of issue. Does the question mean that circulated dimes, being worn, have less numismatic value than new dimes Or do the number simply refer to the number of dimes Clarifying a problem statement is always the first step in problem solving. Why solve the wrong problem well  

A number of standard computer programs easily handle problems of this type such as spreadsheet packages, Matlab, Mathcad, Polymath, and so on as well as symbolic manipulators such as Mathematica, Maple, Derive, etc. Most statistic packages and equation solvers will also solve linear equations and have a simple user interface. 

The solution for this example was obtained via the Excel spreadsheet. The equations can be represented in the matrix notation Ax = b where the arrays are 

To solve for the elements inx, you need to form the matrices (arrays) A, b, and x calculate the inverse of A, A 1 and multiply as follows  

Excel carries out these operations transparently for you. The first screen shows the worksheet with the data entered along with the dark cells reserved for the elements of the inverse matrix A 1. The next figure show the elements of the inverse matrix A-1 and the solution (in column H). 

---

Progress in mean of modelisation and inverse problem solving [1] let us hope to dispose soon of these tools for flaws 3D imaging in Non Destructive Control with eddy current sensors. This will achieve a real improvement of the actual methods, mainly based upon signature analysis. But the actual eddy current probes used for steam generators tubes inspection in nuclear industry do not produce the adequate measurements and/or are not modelisable. 

Special Inspection Problems Solved by Means of Matched Ultrasonic Probe Design. 

Inference engine The inference engine represents the central problem-solving subsystem. It contains strategies for using the information contained in the... 

Isoparametric mapping described in Section 1.7 for generating curved and distorted elements is not, in general, relevant to one-dimensional problems. However, the problem solved in this section provides a simple example for the illustration of important aspects of this procedure. Consider a master element as is shown in Figure 2.23. The shape functions associated with this element are... 

NWChem is part of the Molecular Science Software Suite (MS ) which has been recognized by R D Magazine as one of the 100 most technologically signihcant new products and processes of 1999. The other elements of MS are Ecce, which is a problem-solving environment, and ParSoft, which is the underlying libraries and tools for parallel communication and high-performance input/output. All of the MS components are available publicly. 

The Online Learning Center is a comprehensive exclu sive website that provides a wealth of electronic re sources for instructors and students alike For students the OLC features tutorial problem solving strategies and assessment exercises for every chapter m the book that were developed by Ian Hunt and Rick Spinney from the University of Calgary You can also access the Essential Student Partner from the OLC Log on at WWW mhhe com/carey... 

Problem solving strategies and skills are em phasized throughout Understanding is continu ally reinforced by problems that appear within topic sections For many problems sample solu tions are given... 

Subsection of the analytical approach to problem solving (see Eigure 1.3), of relevance to the selection of a method and the design of an analytical procedure. 

This article provides a basic, step-by-step approach to problem solving in the practice and management of patents and trade secrets. The significance of aggressive patent and trade secret protection to the economic well-being of a business or organization should not be underestimated. Without patents and trade secrets, the marketplace is reduced to competition on the basis of price, which may be very difficult. 

The problem solved by the invention was never recognized before. 

During the 1970s and 1980s, U.S. companies tried to adopt Japanese improvement techniques, but not the philosophy. Thus quahty circles, ie, problem-solving groups of production workers, were initiated (11,12). When this approach failed to achieve anticipated results, it was replaced by other techniques, such as SPG (13), JIT (14), and ZD. This use of contrasting approaches has been summarized (15). 

The Tools of Quality. Quahty assurance also plays an important role in problem solving and process improvement. To do so, QA personnel must be knowledgeable in the many so-called tools of quahty (TOQ) and theh apphcation, so as to guide the efforts of process improvement. Many QA organizations are involved in training employees in these techniques to facihtate quahty improvement. 

Seven of the tools of quahty have been summarized (43). The first tool is a flow chart, used to help understand the organizational flow of a procedure or process. A flow chart should be constmcted with the fiiU participation of the people who do the work. Its principal benefit is to enable teams, such as problem-solving or productivity improvement teams, to reach a common vision of the work flow. Its use enables the improvement effort to begin with this common understanding. Figure 3 contains an example for manufacture of a polymeric material. 

In 1972, the concept of pattern recognition as a general problem solving tool for a broad scope of chemical appHcations was introduced (9,10). 

R. L. Grob and M. A. Kaiser, Environmental Problem Solving Using Gas and Liquid Chromatography, Elsevier, Amsterdam, The Netherlands, 1982. 

Definition / An expert system is a computer program that manipulates large amounts of symboHc knowledge using quaUtative techniques, to solve problems that can otherwise be solved only by expert human problem solvers. Expert systems capture the human problem solver s expertise in the form of domain-specific knowledge and domain-independent problem-solving strategies. 

Representation. Erom a software viewpoint, knowledge-based technology provides ways to represent knowledge and reason with it. The knowledge to be represented includes facts, descriptions, relationships, and problem-solving knowledge. 

---