Optimization multivariate methods

Multivariate chemometric techniques have subsequently broadened the arsenal of tools that can be applied in QSAR. These include, among others. Multivariate ANOVA [9], Simplex optimization (Section 26.2.2), cluster analysis (Chapter 30) and various factor analytic methods such as principal components analysis (Chapter 31), discriminant analysis (Section 33.2.2) and canonical correlation analysis (Section 35.3). An advantage of multivariate methods is that they can be applied in... 

Wu et al. [46] used the approach of an artificial neural network and applied it to drug release from osmotic pump tablets based on several coating parameters. Gabrielsson et al. [47] applied several different multivariate methods for both screening and optimization applied to the general topic of tablet formulation they included principal component analysis and... 

KC Amoldsson, P Kaufmann. Lipid class analysis by normal phase high performance liquid chromatography, development and optimization using multivariate methods. Chromatographia 38 317-324, 1994. 

The advantage of the GA variable selection approach over the univariate approach discussed earlier is that it is a true search for an optimal multivariate regression solution. One disadvantage of the GA method is that one must enter several parameters before it... 

Therefore it is necessary to use methods which allow all variables (substituents) to be investigated simultaneously, i.e. multivariate methods [26], This can be achieved by using design (e.g., factorial design [28] or D-optimal design [29]) in principal properties and evaluating the result by multivariate analysis as shown below. 

Knowledge of multivariate methods is not, however, widely spread in the community of synthesis chemists. Therefore, many new methods are still being investigated through poorly designed experiments and hence, new procedures are not properly optimized. Still, the most common method to carry out "systematic studies" is to consider "one factor at a time", although such an approach was shown by R.A. Fisher to be inappropriate over 60 years ago [1], when several factors are to be considered. 

By analogy with Carlson s" optimization of syntheses by computer-assisted multivariate methods," Fleck" is presently conducting a systematic search for an optimum set of conditions (including the chiral amine component and catalysts) for peptide syntheses by stereoselective 4CC. 

There are a niunber of different experimental design techniques that can be used for medium optimization. Four simple methods that have been used successfully in titer improvement programs are discussed below. These should provide the basis for initial medium-improvement studies that can be carried out in the average laboratory. Other techniques requiring a deeper knowledge of statistics, including simplex optimization, multivariate analysis, and principle-component analysis, have been reviewed (5,6). 

If the loop interactions are not severe, then each single-loop controller can be designed using the techniques described earlier in this section. However, the presence of strong interactions requires that the controllers be detuned to reduce oscillations. There are multivariable control techniques, such as optimal control, that provide frameworks in which to handle the interactions between various inputs and outputs. Optimal control methods also allow one to deal directly with nonlinear system dynamics, rather than the linear model approximation required for the techniques discussed in this section. A thorough presentation of optimal control theory is given by Bryson and Ho (1975). Optimal control is discussed further in Section 9.5. 

In this small-sample set, the f-test does as well as the best multivariate methods. This shows that modeling the correlation structure is not necessarily an advantage if the number of samples is low, or, alternatively, that the true correlation structure has not been captured well enough from the few samples that are available to allow meaningful inference. A definite advantage of the f-test is that it has no tunable parameters and can be applied without further optimization. It should be noted that we do not need to apply multiple-testing corrections in this context since we only use the order of the absolute size of the f-statistics to construct the ROC curves, and not a specific cut-off level a. In other applications, however, this aspect should be taken into accoimt. 

Since many factors will affect experimental results, quite complex experimental designs may be necessary. The choice of the best practical levels of these factors, i.e. the optimization of the experimental conditions, will also require detailed study. These methods, along with other multivariate methods covered in the next chapter, are amongst those given the general term chemometrics. 

This paper describes the development of a novel dynamic predictive and optimal control method for the wet end of a papermaking systems. This part of the system plays an important function in the process in terms of its controllability and potential for optimisation. The wet end process is complicated and the control systems are always multivariable and dynamic in nature. Due to the severe interactions between each variable, general physical and chemistry based modelling techniques cannot be established. As such, feed-forward neural networks are selected as a modelling tool so as to build up a number of non-linear models that link all the variables to the concerned quality outputs and process efficiency. 

This paper has presented the development of a dynamic, predictive and optimal control method for the wet end of a papermaking system. The control of this part of the papermaking process is difficult because of its complex, multivariable and non-linear nature with long time delays. The main objective of this work has been to develop a closed loop control strategy suitable for control of the wet end processes of a paper machine. This includes industrial implementation directed towards achieving optimal control of the wet end of a paper making system. 

The second class of multivariable optimization techniques in principle requires the use of partial derivatives, although finite difference formulas can be substituted for derivatives such techniques are called indirect methods and include the following classes ... 

Nonlinear Programming The most general case for optimization occurs when both the objective function and constraints are nonlinear, a case referred to as nonlinear programming. While the idea behind the search methods used for unconstrained multivariable problems are applicable, the presence of constraints complicates the solution procedure. 

The problem of multivariable optimization is illustrated in Figure 3.4. Search methods used for multivariable optimization can be classified as deterministic and stochastic. 

The method of steepest descent uses only first-order derivatives to determine the search direction. Alternatively, Newton s method for single-variable optimization can be adapted to carry out multivariable optimization, taking advantage of both first- and second-order derivatives to obtain better search directions1. However, second-order derivatives must be evaluated, either analytically or numerically, and multimodal functions can make the method unstable. Therefore, while this method is potentially very powerful, it also has some practical difficulties. 

Many techniques can be used to solve multivariable optimizations. Unfortunately, there is no single best method that applies to every type ofresponse surface. Therefore, I will give a number of different procedures, with the advantages and disadvantages of each one. The reader will then have to decide which one(s) he wishes to use. 

---