This paper is a summarization of the machine learning techniques we applied in the Kaggle competition BNP Paribas Cardif Claims Management [1]. We propose three methods to extract features from the original dataset and four statistical models to analysis the features. Then we compare the performances of different combinations of the features and models and show our ranking in the competition.

We add specifications of location-aware measurements to performance models in a compositional fashion, promoting precision in performance measurement design. Using immediate actions to send control signals between measurement components we are able to obtain more accurate measurements from our stochastic models without disturbing their structure. A software tool processes both the model and the...

application of automatic systems, including discharge measurement technology is one of the effective methods for improving delivery management, water use efficiency, flexibility, and performance of irrigation networks. several aspects such as hydraulic, technical, physical, environmental, economic, social, and managerial, have significant impact on these systems selection. due to diversity of t...

Background: Exposures are occasionally only available in an aggregated form, such as an average physical exposure for each job title. When quantifying the relationship between an individual level outcome and a group level exposure, the variance estimates are contaminated with Berkson measurement error. If the aggregated exposures come from a separate sample, such as a job exposure matrix, the a...

    Six sigma,   Rolled throughput yield, Organizational performance Six Sigma is a well- established approach to improve the capability of business processes in order to gain satisfaction of customers. The performance assessment of a given process is essential to some phases of six sigma methodology. So far, different indicators are used to demonstrate the performance of a process, while many ...

in recent years, the balanced score card (bsc) has been the focus of considerable methodology for strategic cost management in management accounting area since it was originally proposed by kaplan ,r.s. and norton, d.p. in 1992. bsc is an approach to performance measurement based on both financial and nonfinancial information from four perspectives which are called financial, customer, internal...

The Laplace approximation is applied to measurement error models when the likelihood function involves high dimensional integrals. The cases considered are generalized linear models with multiple covariates measured with error and generalized linear mixed models with measurement error in the covariates. The asymptotic order of the approximation and the asymptotic properties of the Laplace-based...

In this paper, we investigated model-based available bandwidth measurement via the use of an active probing stream. An end-to-end measurement framework, which summarizes the current research efforts and highlights insights for measurement practice, is proposed. This framework helps the design of more efficient probing methodologies and enhances understanding of network performance. Of particula...