### TABLE II MEMBERSHIPSOF VARIABLESAND LOCAL PRINCIPALCOMPONENTVECTORSWITH STANDARDFUZZIFICATION(ARTIFICIALDATA SET) probabilistic constraint possibilistic constraint

### TABLE III MEMBERSHIPSOF VARIABLESAND LOCAL PRINCIPALCOMPONENTVECTORSWITH ENTROPYREGULARIZATION(ARTIFICIALDATA SET) probabilistic constraint possibilistic constraint

### Table 1: Evaluation on five models. The accuracy for SLIF and Document classification and F1- measure for named entity extraction are reported. We compared stacked graphical model to a local model, another relational model, and its probabilistic ceiling. The local models for SLIF and Document classifi- cation are MaxEnt models, and the local models for named entity extraction are CRFs. The competitive relational models for SLIF and Document classification are RDN models, and the competitive relational models for named entity extraction are stacked-CRFs.

2007

"... In PAGE 8: ...odel. For the name extraction, the local model is a CRF model. The third and fourth models are stacked graphical models. The fifth model is a probabilistic upper-bound (noted as ceiling model in Table1 ) for the stacked graphical model, i.e.... ..."

Cited by 2

### Table 1: Evaluation on flve models. The accuracy for \SLIF quot; and \Document classiflcation quot; and F1-measure for named entity extraction are reported. We compared stacked graphical model to a local model, another relational model, and its probabilistic ceiling. The local models for \SLIF quot; and \Document classiflcation quot; are MaxEnt models, and the local models for \named entity extraction quot; are CRFs. The competitive relational models for \SLIF quot; and \Document classiflcation quot; are RDN models, and the competitive relational models for \named entity extraction quot; are stacked sequential CRFs.

### Table 1. The incremental localization algorithm.

"... In PAGE 15: ....4. Algorithmic complexity The complexity of the learning and the performance methods must be analyzed separately. The localization algorithm described in Table1 must be executed in real time, while the robot is in operation, whereas the learning algorithm described in Table 2 can be run offline. Our primary concern in the analysis is time complexity.... In PAGE 15: ....4.1. Localization The complexity of probabilistic localization ( Table1 ) depends on the representation of P(fj ) and Bel( ). In the worst case, processing a single sensor reading requires O(Kn + nW) time, where K is the trainingset size, n is the number of networks and W is the number of weights and biases in each neural network.... ..."

### Table 2. Performance of positional baseline, decision-based, and probabilistic systems (precision, recall, and F-measure). Probabilistic systems

2003

"... In PAGE 6: ... Even though these sets are inde- pendent, both contain sample essays from all prompt topics. Table2 compares the overall performance of the decision- and probabilistic- based systems to the positional baseline. Three of the four systems (decision-based, proba- bilistic-local, and probabilistic-global) signif- icantly outperform the baseline.... In PAGE 6: ...sed C5.0 for our voting models. Table 3 compares the positional baseline system, the best single system (that is, the decision-based system), and a voting system. For the single system, the results in Table 3 represent the same runs used in Table2 for the decision-based system Using the 10-fold cross-validation, the voting algorithm outperforms the baseline algorithm and the single system at both the category and overall system levels. Topic independence In a classroom, teachers can give students writing assignments on any topic.... ..."

Cited by 19

### Table 1. The incremental localization algorithm.

1996

"... In PAGE 15: ....4. Algorithmic complexity The complexity of the learning and the performance methods must be analyzed separately. The localization algorithm described in Table1 must be executed in real time, while the robot is in operation, whereas the learning algorithm described in Table 2 can be run offline. Our primary concern in the analysis is time complexity.... In PAGE 15: ....4.1. Localization The complexity of probabilistic localization ( Table1 ) depends on the representation of a4 a6a12 a23 a5 a1a15a14 and a7a9a8 a10 a12 a1a15a14 . In the worst case, processing a single sensor reading requires a0 a12 a10 a1 a3 a1 a2a1 a14 time, where a10 is the trainingset size, a1 is the number of networks and a1 is the number of weights and biases in each neural network.... ..."

### Table 2: Worst case, average case, vs. probabilistic case

1997

"... In PAGE 4: ... This is because in each iteration the ET method finds the local optimal place; however, scheduling nodes to these positions does not always result in the global optimal schedule length. In Table2 , based on the system that has 2 adders and 1 mul- tiplier, we present the comparison results obtained from apply- ing list scheduling, traditional rotation scheduling, probabilistic rotation scheduling using TS, and traditional rotation schedul- ing considering expected computation times, to the benchmarks. Columns L and R show the schedule length obtained from applying list scheduling and traditional rotation scheduling using TS when considering the worst case scenario.... ..."

Cited by 3

### Table 3 lists the comparison between various range free schemes. Tian He et al. experiment with several parameters such as node density (ND), anchors heard (AH), anchor to node range ratio (ANR), anchor percentage (AP), degree of irregularity (DOI), GPS error and placement of node and anchors. Recently research on localization is focused on incorporating the mobility model. Although mobility makes the analysis more difficult, more accuracy is obtained. In [15], Lingxuan Hu etal. use a sequential Monte Carlo Localization method and argues that they exploited mobility to improve accuracy and preci- sion of localization. Probabilistic techniques, such as Markov modeling, Kalman filtering and Bayesian analysis can also be used to determine the absolute loca- tion of a mobile node [18].

2005

"... In PAGE 10: ... Table3 . Comparison of range free schemes Table 4 gives a global view of localization techniques classified by achievable accuracy and the type of location estimation used for various technologies.... ..."

Cited by 3