Search Results for "using"

The CausalImpact repository houses an R package that implements causal inference in time series using Bayesian structural time series models. Its goal is to estimate the effect of an intervention (e.g. a marketing campaign, policy change) on a time series outcome by predicting what would have happened in a counterfactual “no intervention” world. The package requires as input a response time series plus one or more control (covariate) time series that are assumed unaffected by the intervention, and it divides the time horizon into “pre-intervention” and “post-intervention” periods. ...

...Its primary focus in on the interpolation / regression technique known as kriging, which is very closely related to Splines and Radial Basis Functions, and can be interpreted as a non-parametric Bayesian method using a Gaussian Process (GP) prior. The STK also provides tools for the sequential and non-sequential design of experiments. Even though it is, currently, mostly geared towards the Design and Analysis of Computer Experiments (DACE), the STK can be useful for other applications areas (such as Geostatistics, Machine Learning, Non-parametric Regression, etc.).

The new version of MinimPy (MinimPy2) is available for download at: https://osdn.net/dl/minimpy2/MinimPy2.zip Please send your feedback about this new version to mahmood.saghaei@gmail.com (include MinimPy2 in the subject) ============================================================ MinimPy is a desktop application program for sequential allocation of subjects to treatment groups in clinical trials by using the method of minimization. Comprehensive reference help is available at http://minimpy.sourceforge.net MinimPy has been fully described in the following article: Saghaei, M. and Saghaei, S. (2011) Implementation of an open-source customizable minimization program for allocation of patients to parallel groups in clinical trials. ...

An R package for metrology. metRology provides classes and calculation and plotting functions for metrology applications, including measurement uncertainty estimation and (eventually) inter-laboratory metrology comparison studies.

...It is an alternative to randomization, with the advantage of balancing arms of trial with respect to preselected prognostic factors. All aspects of minimization procedure including treatments, factors, minimization protocole, and including a preload can be customized using QMinim.

A Matlab software routine to perform Principal Component Analysis using Covariance, Correlation or Comedian as the criterion. Though, initially developed for experiments related to fretting wear but can be effectively used to interpret experimental data from any field. The attached files contain source code as well as a sample MATLAB (.mat) data file of 13 variables. It could be replaced to the data file of your choice.

...The framework can easily be extended to other model types, e.g. growth/no-growth boundary models. PMM-Lab has been initiated and provided by the Federal Institute for Risk Assessment - BfR (Berlin, Germany). The software is in Beta status. Before using the software you have to read and accept the license and disclaimer (https://sourceforge.net/p/pmmlab/wiki/Disclaimer/). If you do not agree, do not use this software.

Libagf is a machine learning library that includes adaptive kernel density estimators using Gaussian kernels and k-nearest neighbours. Operations include statistical classification, interpolation/non-linear regression and pdf estimation. For statistical classification there is a borders training feature for creating fast and general pre-trained models that nonetheless return the conditional probabilities. Libagf also includes clustering algorithms as well as comparison and validation routines. ...

QPNet (Quick Petri Net) is a fast Petri Nets' emulator, created by students of Moscow State Technical University of Radio Engineering, Electronics and Automation. It brings benefits for the educational process, while being under further development and extension. QPNet это быстрый эмулятор сетей Петри, созданный студентами МИРЭА. Он применяется в учебном процессе, а так же продолжает разрабатываться и эволюционировать.

...Mixed models partition mortality into exogenous and endogenous components, so that the intrinsic survivorship can be estimated without the interference from extrinsic noise. DeDAY supports both interval-censored data and exact event-time data. Using MLE (Maximum Likelihood Estimate), DeDAY fits statistic model to the data. DeDAY also calculates the variances and the multi-dimensional confidence limits of model parameters. DeDAY is free for academic users.

This is a small R package for doing Weibull-based reliability analysis. This R package is now obsolete and has been superseded by 'project Abernethy' on http://r-forge.r-project.org/projects/abernethy/.

An R package implementation of a consensus clustering methodology. This package allows users to perform re-sampling statistics based clustering using multiple clustering algorithms to assess the robustness of both clusters and members of clusters.

...The considered implemented parametric and non-parametric Monte Carlo likelihood methods were chosen for their ability to treat both one-dimensional and multivariate SDE systems, although the quality of the inferential results can't match those obtained using more advanced techniques. Nevertheless the toolbox capabilities to simulate numerical solutions of SDE systems are still valid and can serve as a useful starting point to those willing to simulate stochastic dynamical models easily.

Statastizard(c)2012 Nick Emblow is a program designed to guide its user into the right statistical test, perform the test, and report the results in a meaningful way. The idea of this is that the user won't actually need to know that they are using a mann-whitney U over a t-test, for example - it just tests for normality and homoscedastic of variance, and decides which test to use, then reports the results. Why? Because time after time, people come to me with some test that they've done that they think "proves" something - when in reality they've either not done a comparative test to start with (ex. a correlation), failed to test for confounding factors (such as normality etc) and as a result, failed to appropriately identify even the right test to do.