Modeling Sample Clauses

Modeling. With respect to the Modeling, Critical shall retain ownership of the analytical process. CRC shall retain ownership of all data provided for the Modeling and all results of the application of the analytical process to the data. Critical shall not, without prior written permission of CRC, transfer, disclose or otherwise provide the data or results of the Modeling to any person outside of Critical. Critical agrees that it shall thoroughly safeguard the confidentiality of the data in the Modeling results, and in no event shall it be to a lesser extent than Critical safeguards its own proprietary information. Critical agrees that access to such data and the Modeling results will be given only to employees of Critical who require access in the course of Critical's business, and such employees will be informed of the confidential nature thereof and shall be required to observe provisions of confidence as set forth herein. 1. Within seven days following termination or expiration of this Agreement, Critical shall return all data provided by CRC for the Modeling and all Modeling results. Thereafter, within said seven day period, Critical shall destroy all copies of the Modeling Data provided by CRC and the Modeling results which Critical has in its possession.

Modeling. Seller shall provide all of the data to allow the modeling of the generators, transformers and control systems within the Facility. Seller shall validate or update the modeling data as requested by Company.

Modeling. The CAISO and Bonneville will update, improve, and maintain the modeling of generation and transmission topology to adequately reflect the expected real-time system impacts on Bonneville’s transmission system and the EIM Area based on the data and information shared pursuant to this Agreement and data and information shared through NERC and WECC reliability standards or other regulations, Peak Reliability’s Universal Data Sharing Agreement or its successor, other applicable Peak Reliability policies or methodologies, and other agreements between the Parties.

Modeling a. Resource Provider shall provide PREPA with a PSS/E model for the Facility for approval no later than the Agreement Date. b. Resource Provider agrees to keep the PSS/E mathematical models current with the future versions of the PSS/E program, and shall provide updated PSS/E mathematical models to PREPA not later than [ninety (90)] Days after PREPA notifies Resource Provider of a PSS/E version upgrade if such upgrade results in software incompatibility with PREPA’s system. Resource Provider shall submit to PREPA a report from Siemens PTI or another third-party engineering consultant that validates and certifies the PSS/E mathematical model as accurate, including the subsequent revisions performed to keep the mathematical model current with the future version of the PSS/E program. PREPA shall bear all costs incurred by Resource Provider in excess of the Modification Limit in connection with changes to the PSS/E mathematical model that result from modification or expansion of the MTRs or PREPA’s requirements for protective devices in the Interconnection Facilities as per Section 4.2 (Modifications).

Modeling. The HAM-D scores at each visit were used as the random variable of interest. Two questions were considered in this research. The first question is if the combination of medication and therapy proves to be more effective than medication alone. The second question is if the addition of therapy onto medication changes the trajectory of depressive symptoms, i.e. does it increase the rate of remission. The second question is the topic of this thesis and also of interest to psychiatric researchers. In our models, we consider time, group, and their interaction as the primary predictors. There are seven time points in phase one and seven time points in phase two, with an overlap where the last visit of phase one is the first visit of phase two. Group has three levels throughout both phases: CBASP and medication, BSP and medication, and medication alone. The difference between the models is how time is treated. The first four models will only look at phase two and the last model will consider the change in slope from phase one to phase two in order to compare the treatment groups. The first two models that were run on phase two data were a repeated measures ANOVA. In order to use a repeated measures ANOVA, the data needs to be made complete. For model one, all patients with missing data were deleted to make the complete case data set which was then used to run the analysis. For model two, the LOCF method was used to complete the data. Although this is a biased analysis, it is still often used and still currently approved by the FDA. The third model run on the phase two data was a MMRM model using a random intercept; it is less biased because it uses all the data. This approach is the more recent clinical trials standard. The fourth model used on the phase two data was a growth curve analysis using random intercept and random slope. This models time as a continuous variable and specifically tests the differences in slopes between groups as opposed to just comparing pre-post. Phase one and phase two data were then used to answer the second question of whether the addition of psychotherapy to medication increases the rate of the improvement in the HAMD-D score. For this model, a piece-wise growth curve analysis was fit with a knot at twelve weeks using random intercept, random phase one slope, and random difference between phase two slope and phase one slope. Week twelve is the last visit of phase one and was considered baseline for phase two. This model allows the co...

Modeling a. Section 3.02(i) of the Agreement is hereby deleted in its entirety and replaced with the following: (i) [***] If Comscore requests Charter’s approval for an exception to this Section 3.02(i), Charter shall provide final approval or explanation for non-approval within [***] (e-mail acceptable) of receipt of such request, such approval not to be unreasonably withheld. b. Section 3.03 of the Agreement is hereby amended by inserting the following at the end: [***]

Modeling. 4.4.1 Pipeline To implement our predictive models we use databricks python spark (pyspark). This is a pretty new programming language. Python allows us to run functions over collections on multiple processors in parallel. Spark allows us to run functions over multiple servers in parallel (Odersky, 2017). This should allow us to analyze massive amounts of tweets in a reasonable time. Working locally on python would be limited because we are tied to a single compute node (computer). To model everything in PySpark, we need to put our data in the right format. First of all we considers null values being the integer 0. Most of the operations will be done on either the categorical values, the continuous values or the label. Therefore, a split-up in a label, the categorical features and the continuous features is the most efficient. The conversion is the easiest with RDD's, therefore we will transform the DataFrame to an RDD and afterwards use a map function to transform each of the rows. After putting it in the right format we convert the RDD back to a DataFrame and we split the base table up in a train and test set. 70% of the data is used for training and 30% is used for testing. The splitting of the data is executed randomly to save computation time. Cross validation is not used because we have limited Databricks resources. Now we have our basetable ready for a Machine Learning Pipeline. So now, we need to construct a pipeline. The pipeline will consist out of five stages:

Modeling. Affected standards: MOD-016-1.1, MOD-017-0.1, MOD-018-0 and MOD-019- 0.1, which will be replaced by MOD-031-1 (effective 7/1/2016); MOD-032-1 R1 (effective 7/1/2015); MOD-032-1 R2, R3, R4 (effective 7/1/2016) HHWP will provide HHWP transmission system load pursuant to the WECC Data Collection Manual and CEC data collection requirements. The CAISO will include this data in its documentation for its Planning Authority Area, developed consistent with its Tariff and BPMs, that identifies the scope and details of the actual and forecast (a) Demand data, (b) Net Energy for Load data, and (c) controllable DSM data to be reported for system modeling and reliability analyses. The CAISO will use the HHWP transmission system load data provided by HHWP as needed to meet its obligations under MOD-016-1.1, MOD- 017-0.1 and MOD-018.0. MOD-019-0.1 is not applicable because there are no HHWP interruptible demands or DCLM load data on the HHWP system.

Modeling. If the groundwater in the production zone is restored to its premining condition on a parameter-by-parameter basis, no further evaluation is required. If there are parameters that exceed the premining conditions, the operator will be required to use appropriate modeling to demonstrate that those parameters above the premining conditions will not degrade the adjacent groundwater to the extent that the groundwater will no longer meet its previous class-of-use and that concentrations of constituents for which MCLs have been established by the U.S. EPA do not exceed those MCLs, or baseline, whichever is higher. If the adjacent groundwater class-of- use will be impacted, the operator will be required to conduct additional restoration. If the modeling indicates the adjacent class-of-use will not be impacted and that concentrations of constituents for which MCLS have been established by the U.S. EPA do not exceed those MCLs, or baseline, whichever is higher, a monitoring program sufficient to verify the model may be required.