Decision tree case studies

Much of what we said about clinical chemistry parameters is also tme for the hematologic measurements made in toxicology studies. Which test to perform should be evaluated by use of a decision tree until one becomes confident as to the most appropriate methods. Keep in mind that sets of values and (in some cases) population distribution vary not only between species, but also between the commonly used strains of species and that control or standard values will drift over the course of only a few years. 

As in every safety study, the solution of heat accumulation problems can be undertaken by applying the two commonly used principles of simplification and worst-case approach, as described in Section 3.4.1. A typical procedure following these principles is illustrated with the example of a decision tree for the assessment of the risks linked with thermal confinement (Figure 13.7). 

Table 6.12 and a decision tree of optimization activities is summarized in Fig. 6.5. Two model low-dose drug candidates, Drugs A and B, serve as case studies for evaluating the process selection and optimization. 

The main rationale for a poststudy inspection is to confirm that the study was carried out to GXP and to the agreed plan, including all set criteria and specifications. The depth of this inspection can vary, ranging from confirmation of audit trail (sample integrity to final results) to evaluation of exceptions, that is, what decisions were made when repeat analyses were carried out, what triggered the repeat, were there appropriate SOPs, and were they followed. If there were exceptions not covered by SOPs, were decisions made objectively and consistently How much decision making was automated, for example, if repeat analyses were carried out, was a decision tree used, and if so, was this automated or manually applied If automated computerized systems are used, any manual intervention should arouse suspicion. In such a case, the level of auditing should be raised. 

All the selection criteria presented in the decision tree in Fig. 15.5-1 are general, and many protein interfaces will only fulfill some of them. In these cases - and also for interfaces that meet all the decision tree criteria - an experimental study of the interface should be carried out before starting drug discovery activities. This experimental validation should enable a good level of confidence to be obtained on the druggability of the selected interface. 

It should be pointed out that this approach can t strictly be used for TTS purposes as acoustic features (e.g. time in seconds) measured from the corpus waveforms were used in addition to features that would be available at run time. Following this initial work, a number of studies have used decision trees [264] [418], and a wide variety of other machine learning algorithms have been applied to the problem including memory based learning [77] [402], Bayesian classifiers [516], support vector machines [87] and neural networks [157]. Similar results are reported in most cases, and it seems that the most important factors in the success of a system are the features used and the quality and quantity of data rather than the particular machine learning algorithm used. 

Table 11.3. The partioning profile of case study chemicals generated from box E of the decision tree analysis.

Aware of the variety of conditions possible, and to help in the task of choosing the appropriate mobile phase for the separation of interest, suppliers of CSPs usually provide tips and decision-tree schemes for method development procedures. Also, diverse authors have published studies on the subject. This is the case for polysaccharide-derived... 

---