Defining the design space

Given resource constraints in the form of a partial binding / p and a set of resource allocations ai,the design space of a sequencing graph G,(y, E, 6) is defined as follows. 

An important aspect of the design space formulation is that it is a complete charactmzation of the entire set of possible design tradeoffs for a given allocation of resources. This formulation allows partial binding information to be uniformly incorporated, where the partial binding is used to limit the design 

Given a resource type t T and an allocation a(t), the problem is to compute the number of possible bindings of 0(t) operations to a t) resources. 0 t) is the operation set of t. We assume the condition 1 a(f) 0(t) holds because otherwise the allocation is invalid. Without loss of generality, we consider the design space for one resource type only. The extension to support multiple resource types is to form a Cartesian product among the bindings of each resource type. We omit the (t) suffix fitom 0(t) and a i) in the sequel without ambiguity. 

The computation is divided into two phases. The first phase identifies a set of partitions, each of which is a different partition of 0 objects into a ordwed blocks [Aig79]. For example, if we had 3 objects ( 0 = 3) and 2 blocks (a = 2), then there are two possible partitions (2,1) and (1,2). Partition (2,1) means that two objects are in the first block and one object is in the second block partition (1,2) means that one object is in the first block and two objects are in the second block. The number of possible partitions M( 0, a) is defined recursively as follows  

The formula M K, N) computes the number of partitions of K operations into N resources. Three cases are possible  

---

Pharmaceutical formulation and process development should provide sufficient information and knowledge to understand and support establishment of the design space, specifications, and manufacturing control. The design space of QbD is defined as the multidimensional combination and interaction of input variables (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality.6 Because multiple variables in formulation and process can be encountered, it is important to use an effective methodology to define the design space. 

To control the quality of the product, any relationship among the critical raw material attributes, process parameters, and quality attributes for each process should be established dining development. In addition, limits need to be specified for critical process parameters that define the design space where the quality of the product is guaranteed. 

Six different transients were analyzed with RELAP5-3D. Transients based on normal and casualty conditions were chosen to evaluate the response of the reactor and define the design space within which the reactor and the remainder of the plant would need to perform. Each of those transients is discussed in the following sections. 

The recent FDA initiative on Process Analytical Technology (PAT) attempts to define a design space for unit operations such as compaction... 

Evaluation of candidate compounds is the most critical step in de novo design. It is the duty of the fitness function to decide whether a composed virtual molecule is kept or discarded in deterministic algorithms. One must always be aware that it is the used fitness function that defines the search space for novel molecules. 

The design space is the factor space within this domain defined in terms of the coded variables X,. 

Note that 3 and X4 are qualitative factors and their values are limited to 1. An optimization method that could distinguish between discrete and continuous variables and which would thus only search the space of permitted values of the qualitative, or discrete quantitative variables, would be preferred. It can also be seen that the maximum desirability, on the edge of the square defined by X X2] at 1, is actually just outside the cylindrical experimental domain. The optimum should, strictly speaking, be displaced very slightly to lie on the edge of the design space. 

In setting up the domain the experimenter may define the limits in terms of ratios of one component to another, rather than in terms of proportions (1). These may sometimes be more meaningful as variables for analysis of the data. Unlike all other examples treated in this chapter, the design space is not a simplex, but is generally cubic in form. 

Once the individual limits of each component are known, the design space for the experiment may be defined. For 3 components it may be determined graphically. For more than 3 or 4 components the McLean-Anderson algorithm (6) is used. Experiments must then be selected within the design space. As examples we take the solubility of a drug in a ternary mixture where the constraints give rise to a non-simplex design space and the formulation of a sustained release tablet, with 4 variable components. 

Note that this is a case of a Scheff6 equation being used for a non-simplex experimental domain. The example is evidently a trivial one, as far as defining the domain goes, as the problem of the design space is so easily solved graphically. With 4 or more components it becomes more difficult. 

The concept of data fusion has also been extended to activity landscape modeling." It is well known that activity landscapes will be largely influenced by the choice of the molecular representation that is used to define the chemical space. In an effort to address this issue, multiple structural representations are combined using data fusion to derive a consensus model of activity landscapes and identify consensus activity cliffs [143-145]. Consensus models are designed to prioritize the SAR analysis of activity cliffs and other consistent regions in the activity landscape that are captured by several structure representations. They are not meant to be a means for eliminating data by disregarding, for example, true activity cliffs that are not identified by some structure representations. 

For pharmacy preparations, the process of standardisation of an individual formulation can be seen as defining its design space. This design space should be large enough to enable different pharmacies to prepare a product following the formulation and according to the requirements. Ways to perform that task are ... 

To define the design point, it is necessary to transform the problem posed in Equation 1 from the original random variable space efR (also known as the physical space) to the standard normal space x e 3f . This is, to use a mapping -o- a in order to establish a suitable transformation ... 

A common means to perform such mapping is to use an approximate method like the Nataf s model (Liu and Der Kiureghian 1986). Once the random variables involved in Equation 1 have been expressed into the standard normal space, it is possible to define the design point (x ) using a geometrical or probabilistic interpretation (see, e.g. (Freudenthal 1956)). In the geometrical interpretation, the design point is defined as the realization in the standard normal space which lies on the limit state function... 

---