Learning from

Case-based reasoning. The main advantage of CBR systems for NDT data interpretation is that they can cope with data coming from inspection of varying constructions under varying conditions with various system settings due to their ability to learn from the data classified by the operator. In such situations no reliahle statistical classifier can be designed, and the rule-hased classifiers would be either very inefficient or unpractically complex. 

Our experiences with the software developed within the ANDES project have shown that CBR is a helpful methodology for use in the interpretation of NDT data from field inspections. Because CBR systems can adapt to new situations they can cope with inspection of varying constructions in varying conditions. However, because CBR systems learn from classifications made by the operator this means that they will not be very useful for completely automatic interpretation. Fortunately, most of the NDT inspection requires the presence of an operator because of the required high reliability. 

B3.1.1.3 WHAT IS LEARNED FROM AN ELECTRONIC STRUCTURE CALCULATION ... 

The secret to success has been to learn from data and from experiments. Chemists have done a series of experiments, have analyzed them, have looked for common features and for those that are different, have developed models that made it po.ssiblc to put these observation.s into a systematic ordering scheme, have made inferences and checked them with new experiments, have then confirmed, rejected, or relined their models, and so on. This process is called inductive learning (Figure 1 -1), a method chemists have employed from the veiy beginnings ol chcmistiy. 

There is, however, another type of learning inductive learning. From a series of observations inferences are made to predict new observations. In order to be able to do this, the observations have to be put into a scheme that allows one to order them, and to recognize the features these observations have in common and the essential features that are different. On the basis of these observations a model of the principles that govern these observations must be built such a model then allows one to make predictions by analogy. 

To recognize reaction classification as an important step in learning from reaction instances... 

Now, chemists have acquired much of their knowledge on chemical reactions by inductive learning from a large set of individual reaction instances. How has this been done And how can we build on these methods and knowledge and perform it in a more systematic manner by algorithmic techniques ... 

The first step in an inductive learning process is always to order the observations to group those objects together that have essential features in common and to separate objects that are distinctly different. Thus, in learning from individual reactions we have to classify reactions - we have to define reaction types that encompass a series of reactions with essential common characteristics. Clearly, the definition of what are essential common features is subjective and thus a variety of different classification schemes have been proposed. 

Two factors matter most in gaming knowledge from data first, the quality of the data and secondly, the method one applies to the data, and by which one learns from them. 

We learn from data. Therefore, the way we prepare the data for the learning process will crucially condition the quality of learning and the reliability of the extracted knowledge. 

Twenty-eight chiral compounds were separated from their enantiomers by HPLC on a teicoplanin chiral stationary phase. Figure 8-12 shows some of the structures contained in the data set. This is a very complex stationary phase and modeling of the possible interactions with the analytes is impracticable. In such a situation, learning from known examples seemed more appropriate, and the chirality code looked quite appealing for representing such data. 

The area of machine learning is thus quite broad, and different people have different notions about the domain of machine learning and what kind of techniques belong to this field. We will meet a similar problem of defining an area and the techniques involved in the field of "data mining , as discussed in Section 9.8. We will use the term "machine learning in this chapter to collect aU the methods that involve learning from data. 

One application of machine learning is that a system uses sample data to build a model which can then be used to analyze subsequent data. Learning from exam-... 

Just like humans, ANNs learn from examples. The examples are delivered as input data. The learning process of an ANN is called training. In the human brain, the synaptic connections, and thus the connections between the neurons. 

Neural networks model the functionality of the brain. They learn from examples, whereby the weights of the neurons are adapted on the basis of training data. 

R. S. Michalski, R. E. Stepp. Learning from Observation Ganceptual Clustering, in Machine Learning An Artificial Intelligence Approach, R.S. Michalski, f.G. CarboneU, T.M. Mitchell (Eds.), Morgan Kauffmatm, San Mateo, GA, 1983, pp. 331-363. 

I-J and G Klebe 1996. What Can We Learn From Molecular Recognition in Protein-Ligand nplexes for the Design of New Drugs Angewandte Chemie Iniemational Edition in English 2588-2614. 

The electron micrographs of Fig. 4.11 are more than mere examples of electron microscopy technique. They are the first occasion we have had to actually look at single crystals of polymers. Although there is a great deal to be learned from studies of single crystals by electron microscopy, we shall limit ourselves to just a few observations ... 

It was thought previously that there were no inborn odor preferences that these are learned from experience. However, studies at the MoneU Center have indicated that flavors consumed by a mother and transmitted into the milk influence the feeding behavior of her infant. When mothers consume gadic, their infants feed longer than when no gadic is consumed (7). 

Some important conclusions can be learned from this simple model. First, it shows that does not depend on polymer molecular weight,... 

Do we recognize that key inventions are often made here Do we learn from our lead end users (39) ... 

Learning Curves It is usual to learn from experience. Consequently, the time taken to produce an article, the number of spoiled batches, the cost per unit of production, etc., tend to decrease with the number of units produced. The relationships are expressed for the ideal case by... 

Control of Plant and Process Modifications Many accidents have occurred because plant or process modifications had unforeseen and unsafe side effects (Sanders, Management of Change in Chemical Plants Learning from Ca.se Histories, Butterworth-Heinemann, 1993). No such modifications shoiild therefore be made until they have been authorized by a professionally quahfied person who has made a systematic attempt to identify and assess the consequences of the proposal, by hazard and operability study or a similar technique. When the modification is complete, the person who authorized it... 

Drogaris, G. 1993. Major Accident Reporting System Lessons Learned from Accidents Notified. Elesevier Science Publishers,B.V., Amsterdam. 

The author is sure that the readers will find ample opportunity to learn from his experience and apply this information to their field of activities. The book aims to provide a bridge between the concept and the application. With this book by his or her side, an engineer should be able to apply better, design better and select better equipment for system needs and ambient conditions. It should prove to be a handy reference to all those in the field of design and application, protection and testing, production, project engineering, project implementation or maintenance, in addition to the sales and purchase of these products. 

A common use of statistics in structural biology is as a tool for deriving predictive distributions of strucmral parameters based on sequence. The simplest of these are predictions of secondary structure and side-chain surface accessibility. Various algorithms that can learn from data and then make predictions have been used to predict secondary structure and surface accessibility, including ordinary statistics [79], infonnation theory [80], neural networks [81-86], and Bayesian methods [87-89]. A disadvantage of some neural network methods is that the parameters of the network sometimes have no physical meaning and are difficult to interpret. 

After the preliminary tests are made on a promising catalyst and some insight gained on the process, it is time to do a kinetic study and model development. The method of a kinetic study can be best explained on an actual industrial problem. Because more can be learned from a failure than from a success, the oxidation of propylene to acrolein is an instructive attempt at process development. (Besides, to get permission to publish a success is more difficult than to solve the problem itself) Some details of the development work follow in narrative form to make the story short and to avoid embarrassing anyone. 

The main lessons learned from the workshop were (with some 5-year old... 

Nicholson, C. E., Heyes, P. F. and Wilson, C. 1993 Common Lessons to be Learned from the Investigations of Failures in a Broad Range of Industries. In Rossmanith, H. P. (ed.). Structural Eailure, Product Liability and Technical Insurance. Elsevier. 

They learn from experience rather than by programming. 

---