### TABLE I THE BEHAVIOR OF THE EXPONENTIAL ARQ FOR DIFFERENT L, p = 0.5, AND Q = 1000 TIMES OF SIMULATION.

### Table 17: Results from applying concept analysis to C programs from the SPEC benchmark (and bash). Objects are functions and attributes are of the form \uses struct t in parameter list or return type quot; plus unique identi cation attributes. \? quot; indicates that the partitioning algorithm never terminated due to exponential behavior.

1997

"... In PAGE 30: ... Instead, we use the two techniques for extending contexts to well- formed contexts described in Section 4. Table17 summarizes the results of applying concept analysis to the well-formed extensions of the contexts used in Table 16, where the extensions are produced by adding unique identi cation attributes. Table 18 summarizes the results of applying concept analysis to the complemented extensions of the contexts used in Table 16.... ..."

Cited by 90

### Table 1a: Behavior of Bayes Factor for Exponential (n=1) or Gamma (n gt;1) Likelihood for Given Weight Function (Before Prior Speci cation) Weight

1994

"... In PAGE 17: ...Table1 b: Behavior of Bayes Factor for Normal ( ; 1) Likelihood for Given Weight Function (Before Prior Speci cation) Weight Function e y y y Pivot Point P 0 1 0 Restriction on gt; 0 gt; 0 : Non-neg Integer As ! 0; BF ! 1 1 (Choose M0) 1 As ! P; BF ! 1 1 1 As ! +1 BF ! 1 (Choose M0) 1 (Choose M0) 0 (Choose M1) quot; or # For lt; P # For lt; y ? # For lt; exp(y ? ) N.A quot; or # For gt; P quot; For gt; y ? quot; For gt; exp(y ? ) #... ..."

Cited by 5

### Table 5 shows the effect of increasing global traffic on the circuit establish time for circuit switched approach when the number of disks on each loop is 8 and the request size is 64KB. Evidently, the circuit establish time2 increases exponentially with the percentage of global traffic. To understand this behavior, we counted the number of attempts needed by an initiator to establish a circuit. Graph 2 shows the observed behavior.

"... In PAGE 5: ... Table 6: Average latencies of both approaches with varying percentage of global traffic. Table5... ..."

### Table 3 { larger problem instances

1997

"... In PAGE 4: ... For these examples the exact algorithms fail due to their exponential behavior. The results are given in Table3 . As can been seen PQ is much faster and ad- ditionally obtains smaller average test costs.... ..."

Cited by 2

### Table 2. Checks and verification Here, the problem of space and time complexity of the finite state machines (automata) for recognizing languages arises. In general, the classical regular language operators (concatenation, alternative, repetition) do not introduce any exponential growth of the state space of a parsing finite state automaton. However, behavior protocols employ also the and- parallel, composition, and adjustment operators that introduce exponential complexity of the resulting automata which might lead to the state explosion problem. In fact, the composition and adjustment operators behave better than the and-parallel operator in terms of the required state space as they comprise synchronization of events, thus reducing the interleaving of traces.

2002

Cited by 112

### Table 3: Performance of xed-size queries on uniformly placed exponential-sized objects.

"... In PAGE 7: ....2.3 Fixed-Size Queries on Uniformly Placed Exponential- Sized Objects In this experiment, we applied the query sequences of Group 1 to Data Set 2. The results are given in Table3 . We observe the same behavior as seen in Table 1.... ..."

### Table 2: Results in an exponential domain for fixed maxi- mum error pruning.

1996

"... In PAGE 8: ... In contrast, sliding tolerance pruning results in much more gradual changes in value as the tolerance t increases, making it easier to select good pruning tolerances. The second domain ( Table2 ) is one in which every state has a uniquevalue, leading to worst-case behavior forstruc- tured methods such as SVI. While the problem has a com- pact input description, SVI must (ultimately)perform back- ups for each state, with the additional overhead of tree con- struction.... ..."

Cited by 39

### Table 3: Recovery time constants, (sec), are calculated at four wavelengths.

"... In PAGE 7: ...3 Recovery Essentially, all silica-core bers tend to recover to varying degrees. In Table3 , the time constants are evaluated assuming an exponential behavior in the form of e?t= . There are undoubedly several processes that contribute to recovery phenomena.... ..."

### Table 1b: Behavior of Bayes Factor for Normal ( ; 1) Likelihood for Given Weight Function (Before Prior Speci cation) Weight

1994

"... In PAGE 16: ... JASA 81, 82-86. Table1 a: Behavior of Bayes Factor for Exponential (n=1) or Gamma (n gt;1) Likelihood for Given Weight Function (Before Prior Speci cation) Weight Function e y y y Pivot Point P 0 1 0 Restrictionon gt; 0 gt; 0 : Non-neg Integer As ! ?1; BF ! N.A.... ..."

Cited by 5