Statistical validation

Mayer, David G.; Butler, Douglas N.

doi:10.1016/0304-3800(93)90105-2

Cited by 762 publications

(469 citation statements)

References 17 publications

Supporting

Mentioning

437

Contrasting

Unclassified

Order By: Relevance

“…where x i is the observed number of beads at depth i with a mean ofx i and y i is the modeled probability values at that particular depth (Willmott, 1981;Mayer and Butler, 1993). The value of d will vary between 0 and 1, with a value of 1 indicating perfect model agreement (Willmott, 1981).…”

Section: Statistical Analysesmentioning

confidence: 99%

SeedChaser: Vertical soil tillage distribution model

Spokas

Forcella

Archer

et al. 2007

Computers and Electronics in Agriculture

View full text Add to dashboard Cite

Knowledge of the vertical distribution of surface residues, chemicals, or seeds following tillage operations is of great importance to a wide variety of soil research areas. This paper describes a 1D empirical vertical soil tillage distribution model with 1 cm grid spacing (SeedChaser) that predicts vertical redistribution of weed seeds following user selected (a) sequences of tillage implements, and (b) initial seed distribution values. Results of this model are particularly suited for weed seed emergence modeling. However, the model can be adapted easily to any surface broadcasted agrochemical or incorporated residues. The present model can handle up to 20 passes of user selected sequences of 16 different implements. The majority of prior models examined only the impact of a more limited list of implements at much larger depth intervals, which reduced the predictability of fine-scale vertical movement that may be needed for simulating movements of seeds or chemical granules. SeedChaser consolidates the results from these previous models along with new data on conservation tillage implements into a prediction tool that would have applications both in weed science, as well as other soil research areas. SeedChaser was developed in JAVA and is publicly available via the Internet.

show abstract

Section: Statistical Analysesmentioning

confidence: 99%

SeedChaser: Vertical soil tillage distribution model

Spokas

Forcella

Archer

et al. 2007

Computers and Electronics in Agriculture

View full text Add to dashboard Cite

show abstract

“…Rainfall hyetographs used to drive the model calibration for the oak and pear trees are shown in Figures 5a and 5b. Figures 5c and 5d show the calibration results for both the oak tree (Figure 5c) and the pear tree (Figure 5d) [Mayer and Butler, 1993] for all interception processes was less than 5% for the oak tree, and less than 4% for the pear tree. These differences seemed acceptable considering the measured differences in gross precipitation and that dry canopy surface was assumed when starting simulations.…”

Section: Interception Model Calibration and Sensitivitymentioning

confidence: 99%

A new approach to modeling tree rainfall interception

Xiao¹,

McPherson²,

Ustin³

et al. 2000

J. Geophys. Res.

166

109

View full text Add to dashboard Cite

Abstract. A three-dimensional physically based stochastic model was developed to describe canopy rainfall interception processes at desired spatial and temporal resolutions. Such model development is important to understand these processes because forest canopy interception may exceed 59% of annual precipitation in old growth trees. The model describes the interception process from a single leaf, to a branch segment, and then up to the individual tree level. It takes into account rainfall, meteorology, and canopy architecture factors as explicit variables. Leaf and stem surface roughness, architecture, and geometric shape control both leaf drip and stemflow. Model predictions were evaluated using actual interception data collected for two mature open grown trees, a 9-year-old broadleaf deciduous pear tree (Pyrus calleryana "Bradford" or Callery pear) and an 8-year-old broadleaf evergreen oak tree (Quercus suber or cork oak). When simulating 18 rainfall events for the oak tree and 16 rainfall events for the pear tree, the model over estimated interception loss by 4.5% and 3.0%, respectively, while stemflow was under estimated by 0.8% and 3.3%, and throughfall was under estimated by 3.7% for the oak tree and over estimated by 0.3% for the pear tree. A model sensitivity analysis indicates that canopy surface storage capacity had the greatest influence on interception, and interception losses were sensitive to leaf and stem surface area indices. Among rainfall factors, interception losses relative to gross precipitation were most sensitive •o rainfall amount. Rainfall incident angle had a significant effect on total precipitation intercepting the projected surface area. Stemflow was sensitive to stem segment and leaf zenith angle distributions. Enhanced understanding of interception loss dynamics should lead to improved urban forest ecosystem management.

show abstract

“…Therefore, the gathered test subset is used during the activity "Verify and Evaluate Models" to evaluate the model accuracy when being used with unknown data. Here, error metrics such as the mean absolute error (MAE) and the mean absolute percent error (MAPE) need to be investigated [57], [58], which should be based on absolute differences to prevent the compensation of positive and negative differences. Both measures are recommended, because the MAE quantifies the error and the MAPE generates a relation to observed values.…”

Section: Activities Of the Capacity Management Providermentioning

confidence: 99%

Capacity Management as a Service for Enterprise Standard Software

Müller

Bosse

Turowski

2017

CSIMQ

View full text Add to dashboard Cite

Abstract. Capacity management approaches optimize component utilization from a strong technical perspective. In fact, the quality of involved services is considered implicitly by linking it to resource capacity values. This practice hinders to evaluate design alternatives with respect to given service levels that are expressed in user-centric metrics such as the mean response time for a business transaction. We argue that utilized historical workload traces often contain a variety of performance-related information that allows for the integration of performance prediction techniques through machine learning. Since enterprise applications excessively make use of standard software that is shipped by large software vendors to a wide range of customers, standardized prediction models can be trained and provisioned as part of a capacity management service which we propose in this article. Therefore, we integrate knowledge discovery activities into well-known capacity planning steps, which we adapt to the special characteristics of enterprise applications. Using a realworld example, we demonstrate how prediction models that were trained on a large scale of monitoring data enable cost-efficient measurement-based prediction techniques to be used in early design and redesign phases of planned or running applications. Finally, based on the trained model, we demonstrate how to simulate and analyze future workload scenarios. Using a Pareto approach, we were able to identify cost-effective design alternatives for an enterprise application whose capacity is being managed.

show abstract

Statistical validation

Cited by 762 publications

References 17 publications

SeedChaser: Vertical soil tillage distribution model

SeedChaser: Vertical soil tillage distribution model

A new approach to modeling tree rainfall interception

Capacity Management as a Service for Enterprise Standard Software

Contact Info

Product

Resources

About