1,556 research outputs found

    An extension of a theorem of Darroch and Ratcliff in loglinear models and its application to scaling multidimensional matrices

    No full text
    AbstractLet C=(cij) be an m ×n matrix with real entries. Let b be any nonzero m-vector. Let K = {π:Cπ = b, π ⩾ 0} be bounded. Let x = (x1, x2,…,xn), y = (y1, y2,…,yn) be two nonnegative vectors with y ϵ K and xj = 0 ⇔ yj = 0 for any coordinate j. Then it is shown that there exists a π ϵ K and positive numbers z1, z2,…, zm such that πj = xj∈mi = 1 zciji for all j. Th theorem slightly generalizes a theorem of Darroch and Ratcliff in loglinear models with a completely different proof technique. The proof relies on an extension of a topological theorem of Kronecker to set valued maps and the duality theorem of linear programming. Many theorems in scaling of matrices and multidimensional matrices are direct consequences of this theorem. The main idea is to associate a suitable zero-one matrix of transportation with any multidimensional matrix. Some motivations for scaling applications are also discussed

    Soap Powder

    No full text
    Patent for soap powder that dissolves easily and may be used in all types of water. Soap will make fabrics white and easy to iron

    An investigation of 'minimalist' and 'constructionist' processing strategies in pronoun comprehension

    No full text
    This thesis investigated the use of 'minimal' (gender/number information, linguistic conjunction and thematic role occupancy) and 'non-minimal' information (spatial information and description type - proper name vs. role name i.e. the waiter) in the comprehension of pronouns, (McKoon and Ratcliff, 1992).Fourteen experiments were conducted. Seven experiments measured clause-by-clause reading times; seven measured frequency of reference in sentence continuation tasks. Six reading time experiments used materials pronominally unambiguous on the basis of the gender and number information. Experiments one to twelve also manipulated the spatial location of the characters (together or apart). Experiments five, six, seven, eight, eleven and twelve also manipulated noun phrase conjunction (by the use of 'and'), proposed to be a cue to plural pronoun use (Hielscher and Musseler, 1990; Sanford and Lockhart 1991). Experiment thirteen was a reading time task using pronominally ambiguous sentences. Characters' thematic roles and description type were manipulated. Experiment fourteen was a sentence continuation task version of experiment thirteen. For experiments one to twelve, the predictions were that subjects making use of a constructionist processing strategy would read plural references faster when characters were described as being together rather titan apart. In continuation tasks, it was predicted that subjects would make more plural references to characters described as being together rather than apart. In experiment thirteen it was predicted that subjects making use of a minimalist strategy would read references faster depending on the character's thematic role occupancy rather than on description type. In experiment fourteen it was predicted that subjects using a minimalist strategy would make more references to characters on the basis of thematic role occupancy rather than description type. The results did not consistently support either the minimalist or constructionist hypotheses. Subjects appeared instead to be making use of different strategies as a function of task demands. This interpretation is in line with work by Garnham et al (1992), McKoon and Ratcliff (1992), and Oakhill et al (1989)

    Particle identification at an asymmetric B Factory

    No full text
    Particle identification systems are an important component of any detector at a high-luminosity, asymmetric B Factory. In particular, excellent hadron identification is required to probe CP violation in B{sup 0} decays to CP eigenstates. The particle identification systems discussed below also provide help in separating leptons from hadrons at low momenta. We begin this chapter with a discussion of the physics motivation for providing particle identification, the inherent limitations due to interactions and decays in flight, and the requirements for hermiticity and angular coverage. A special feature of an asymmetric B Factory is the resulting asymmetry in the momentum distribution as a function of polar angle; this will also be quantified and discussed. In the next section the three primary candidates, time-of-flight (TOF), energy loss (dE/dx), and Cerenkov counters, both ring-imaging and threshold, will be briefly described and evaluated. Following this, one of the candidates, a long-drift Cerenkov ring-imaging device, is described in detail to provide a reference design. Design considerations for a fast RICH are then described. A detailed discussion of aerogel threshold counter designs and associated R D conclude the chapter. 56 refs., 64 figs., 13 tabs

    Measurement of the charged multiplicity of Z0 ---> b anti-b events

    No full text
    Using an impact parameter tag to select an enriched sample of Z0 --> bbBAR events, we have measured the difference between the average charged multiplicity of Z0 --> bbBAR and Z0 --> hadrons to be n(b)BAR-n(had)BAR=2.24 +/- 0.30(stat) +/- 0.33(syst) tracks per event. From this, we have derived n(b)BAR-n(uds)BAR=3.31 +/- 0.41 +/- 0.79. Comparing this measurement with those at lower center-of-mass energies, we find no evidence that n(b)BAR-n(uds)BAR depends on energy. This result is in agreement with a precise prediction of perturbative QCD, and supports the notion that QCD remains asymptotically free down to the scale M(b)2
    corecore