In this module, students will learn how the Least Squares fit statistic can be expressed as a likelihood.
Category Archives: R
Model validation methods
In this module, students will become familiar with the importance of, and methods for, model validation.
Population standardized regression methods
In this module, students will become familiar with population standardized regression methods when working with data that is expressed as per-capita rates
Logistic (Binomial) regression
In this module, students will become familiar with logistic (Binomial) regression for data that either consists of 1’s and 0’s (“yes” and “no”), or fractions that represent the number of successes out of n trials. We focus on the R glm() method for logistic linear regression. Continue reading
Poisson regression
In this module, students will become familiar with Poisson regression for count data. We focus on the R glm() method for linear regression, and then describe the R optim() method that can be used for non-linear models. Continue reading
Students t- and z-tests of sample means, and ANOVA to compare multiple means
In this module we will discuss how to conduct one-sample and two-sample Students t-tests of sample means when the variance of the sample is unknown, testing the equality of the means of several samples, and z-test of sample means when the variance is known.
Contents:
- Students t-test of the mean of one sample
- Example of Students t-test of the mean of one sample
- Students t-test comparing the means of two samples
- Example of Students t-test comparing the means of two samples
- Limitations of the Students t-test
- Testing for equality of more than two means (ANOVA)
- One and two sample Z-tests
Protected: AML 612 Spring 2018: project prospectus list and scoring rubrics
Introduction to hypothesis testing
[In this course we will be learning how to formulate figure-of-merit statistics that can help to answer research questions like “Is quantity A significantly greater/less than quantity B?”, or “Does quantity X appear to be significantly related to quantity Y?”. As we are about to discuss, statistics that can be used to answer these types of questions do so using the underlying probability distribution to the statistic. Every statistic used for hypothesis testing has an underlying probability distribution.]
Bringing together disparate sources of data
In this module, we will discuss how to “mesh” two or more data sets together in order to answer a research question.
Sunspot activity and influenza pandemics: a statistical assessment of the purported association
The R code associated with the analysis presented in the paper Sunspot activity and influenza pandemics: a statistical assessment of the purported association is in sunspot_analysis.R
The file reads in the data files summary_pandemic_data.txt, and sunspot_wolf_and_group_1700_to_2014.txt
The R script produces the following plot, shown in the paper,
Incorporating prior belief of parameter estimates into model fits to data
For our models, we often obtain estimates of some of the parameters from prior studies in the literature. For example, for most diseases, the incubation and infectious periods are usually estimated from observational studies. For animal populations, the birth rates are also usually estimated from observational studies.
However, these parameter estimates in the literature are not a fixed exact value, and have uncertainties associated with them. Those uncertainties, when properly taken into account in our fits using the graphical Monte Carlo method, will feed into the uncertainties on the parameters we are trying to fit for. In this module, we will discuss how to incorporate this “prior-belief” of known parameters and their uncertainties into our graphical Monte Carlo fits.
A (sometimes) better method for estimation of confidence intervals compared to the fmin+1/2 method: the weighted mean method
In this past module, we talked about the “fmin+1/2” method that can be used to easily estimate one standard deviation confidence intervals on parameter estimates when using the graphical Monte Carlo method to fit our model parameters to data by minimizing a negative log likelihood goodness of fit statistic. In this module, we will discuss an alternate method to the fmin+1/2 method for estimating parameter uncertainties, the works for many cases, and also provides an estimate of the covariance matrix of the parameter estimates (something that is very difficult to do with the fmin+1/2 method).
Protected: Temporal and geospatial patterns in bomb threats to US Jewish community centers: Jan to Mar 2017
temp
[CP_CALCULATED_FIELDS][CP_CALCULATED_FIELDS_VAR name=””]
Protected: AML 612 spring 2017: project prospectuses
another test
JavaScript program to calculate multiplication and division of two numbers
The Result is :
temp calculations
Example of applying Least Squares, Poisson likelihood, and Negative Binomial likelihood fitting methods to count data
[This module compares the results of fitting a model to data when Least Squares, Poisson and Negative Binomial likelihood statistics are optimized]
Monte Carlo methods for assessing uncertainty in model estimates due to uncertainties in parameters
[This module presents Monte Carlo stochastic methods that can be used to assess uncertainty in model estimates due to uncertainty in one or more parameters]
Combining stochastic modelling methods: MCMC and SDE’s
[In past modules, we have discussed Markov Chain Monte Carlo methods and Stochastic Differential Equations with Gaussian noise for stochastic modelling of compartmental models. In this module, we will describe how combing the two has the potential to simultaneously optimize computational efficiency and accuracy]
Continue reading