### Table 2: Results for general zero-one problems.

1996

"... In PAGE 17: ... This was one of the early computational breakthroughs in combinatorial optimization, as most of the problems were considered not amenable to exact solution within reasonable time. Table2 gives a summary of the computational results. The valid inequalities were generated and added in the root node of the branch-and-bound tree only.... In PAGE 17: ... For more details about preprocessing we refer to Part II of this article. In Table2 vars, constr., and ineq.... ..."

Cited by 4

### Table XVIII. Optimization scenarios and models. Scenario Source Destination Channel Optimization Model 1 Single Single Single Zero-One KP 2 Single Single Many MKP

2005

### Table 3: Yeast protein data zero-one loss results.

2004

"... In PAGE 4: ...004. We obtained similar results with experiments on the yeast protein data, shown in Table3 . CI resulted in the lowest zero- one loss and its loss was significantly different than the zero- one loss of all other models.... ..."

Cited by 43

### Table 9: Results for general zero-one problems.

1996

Cited by 4

### Table 9: Results for general zero-one problems. 5 Alternative Techniques In the last two decades there has been a remarkable development in polyhedral techniques leading to an increase in the size of many combinatorial problems that can be solved by a factor hundred. Most of the computational successes have occurred for zero-one combinatorial problems where the polytope is de ned once the dimension is given, such as the traveling salesman problem. For more complex combinatorial optimization problems, and for general integer programming problems less progress has been made. Here we shall give a brief overview of other available solution techniques. If the number of variables is large compared to the number of constraints column gen- eration may in many cases be a good alternative. It can be viewed as a dual approach to polyhedral techniques in the sense that one aims at generating the extreme points of conv(S) rather than its facets. Instead of solving a separation problem to generate a violated inequality 35

1996

Cited by 4

### TABLE 1 NUMBER REPRESENTATIONS FOR SEQUENCES OF ZEROES AND ONES

### Table 1: The linear spans of the zero-one sequences from monomial hyperovals.

"... In PAGE 6: ...omputer results when m is small (with the help of R. M. Wilson). The method we used in the calculation is to nd a trace representation for the sequence in question. We list the results in Table1 below. Note that here the sequences are 0,1 sequences which are de ned as follows.... In PAGE 7: ...Table1 lists the sizes of the eld. The next three columns list the linear spans of the 0,1 sequences from the Segre hyperoval, and the two Glynn hyperovals respectively.... In PAGE 7: ...The next three columns list the linear spans of the 0,1 sequences from the Segre hyperoval, and the two Glynn hyperovals respectively. The linear spans in the rst column of Table1 are particularly interesting. The second factors of the linear spans in the rst column satisfy a recursive relation like that of Fibonacci numbers.... ..."

Cited by 1

### Table 1: The linear spans of the zero-one sequences from monomial hyperovals.

"... In PAGE 6: ...omputer results when m is small (with the help of R. M. Wilson). The method we used in the calculation is to find a trace representation for the sequence in question. We list the results in Table1 below. Note that here the sequences are 0,1 sequences which are defined as follows.... In PAGE 7: ...Table1 lists the sizes of the field. The next three columns list the linear spans of the 0,1 sequences from the Segre hyperoval, and the two Glynn hyperovals respectively.... In PAGE 7: ...The next three columns list the linear spans of the 0,1 sequences from the Segre hyperoval, and the two Glynn hyperovals respectively. The linear spans in the first column of Table1 are particularly interesting. The second factors of the linear spans in the first column satisfy a recursive relation like that of Fibonacci numbers.... ..."

Cited by 1

### Table I. The weights for each edge of G are computed via TagLink string metric, [Camacho and Salhi 2006]. In each case, the quality is computed for 31 range values of parameter k, between zero and one. The results are displayed in Fig. 1.

### Table 4: Zero, One and Two Bumps: Proportion of times H0 is rejected at level based on 500 simulations of Yi = m(tija; b) + quot;i, i = 1; : : : ; 101, quot;i N(0; (0:005)2).

"... In PAGE 11: ... The case with a = 0:48 and b = 0:15 continues to pose a problem for CriSP. Insert Table4 about here. 6 Analysis of the Growth Data We consider growth data from a longitudinal study conducted in Berkeley, California.... ..."