A step towards unification of syntactic and statistical pattern recognition

The problem of pattern recognition is discussed in terms of single-entity representation versus multiple-entity representation. A combined syntactic-semantic approach based on attributed grammars is suggested. Syntax-semantics tradeoff in pattern representation is demonstrated. This approach is intended to be an initial step toward unification of syntactic and statistical approaches to pattern recognition.     

Variation in light scattering and water-holding capacity along the porcine Longissimus dorsi muscle

The meat quality of M. longissimus muscle was evaluated in 851 pigs by using the Fibre Optic Probe (FOP) at three sites in the muscle and in both halves of the carcass. A systematic difference between sites was found, with the lowest light scattering (indicating the best meat quality) in the mid-part of the muscle and higher light scattering in the anterior and posterior parts. A non-systematic variation was also observed, with the opposite pattern in some animals, even though it was not frequent. A negative influence of the shackling was found, yielding higher FOP values in the shackled half.

Drip loss measurements in the Longissimus muscle, taken from another 100 pig carcasses, were evaluated using three methods. Drip loss, too, showed a considerable variation along the Longissimus muscle, with minimum losses around the last rib. Repeatability estimates, calculated from two non-consecutive pieces of the Longissimus dorsi muscle from each carcass, varied from 0·4 when keeping samples vacuum packed for 2 days, to 0·5 when the samples were either kept in plain plastic bags for 2 days or in a meat container with a squared inset for 1 or 2 days.     

Comparison of M. longissimus dorsi pigment concentration from implanted and control Angus and Limousin bulls and steers

M. longissimus dorsi total pigment concentration, visual color score and 24-h pH values were evaluated in a 2(3) factorial design that included bulls versus steers, Zeranol implants versus control and Angus versus Limousin comparisons. Bulls had greater total pigment concentration than steers (3.25 versus 2·90 mg/g; P < 0·01) and darker colored lean (P < 0·01). Twenty-four hour pH values did not differ between bulls and steers. Zeranol implanting and breed had no effect on total pigment concentration or visual color score; however, Limousin had higher (P < 0·05) 24-h pH values than Angus. The initial slaughter group (N = 10; average age = 256 days) had 34% less total pigment than the final slaughter group (N = 48; average age = 458 days). The correlation between visual color score and total pigment concentration was -0·65. These results indicate that the darker colored lean from bulls is due in part to an increase in pigment concentration.     

Some techniques for solving linear equation systems with guarantee

Methods are described for solving a system of linear equations with error bounds. Rectangular and spherical intervals ofR ⁿ are used combined. The objective is to get guaranteed accuracy with a minimal effort of computing time.