Computational statistics

Statistics would definitely have been a subject in your high school or in your graduation. Statistics is a science which is used for the purpose of analyzing data. This means that you are extracting knowledge through data. Data is used for the purpose of making decisions. Statistics will be accommodating within itself all data and you can yourself view that randomness in data in the different statements that are used for decision making. Computational statistics which is also referred to as the statistical computing provides an interface between the computer science and statistics. This is an application of scientific computing or commonly known as computational science and this is specific to mathematical statistics. You can make use of both terms interchangeably. There was a proposal that was made by the Carlo Lauro -former president of statistical computing international association that both of these terms should not be used interchangeably. Both of them have a difference in them.

Statistical computing means when you are applying the computer science to the statistics where as computational statistics makes use of statistical methods for the designing of algorithms on computers. Examples include- simulation and bootstrap. This is also used for the purpose of coping with analytical intractable problems. This science of statistics also makes use of ‘resampling’ methods as well as intensive statistics methods. Statistics computing is one of the areas where you will be require doing visualization of statistics by making use of the statistics intensive methods.

This science of statistics is created by the mathematical theory and also makes use of methods of Monte Carlo. It focuses on different exploratory methods. Research in this field will be involving the development of computationally intensive and visualization methods for the purpose of mining non-homogenous, large and multi-dimensional sets of data. This will be used for the purpose of discovering knowledge in data. You should always remember that statistics is incomplete without probability models. One of the most important activities in statistics computing is of evaluation and model building.

There are numbers of techniques available through which you can do the discovering of structures in data. Mainly, these are visually or exploratory and these will be making use of the clustering, classification and density estimation. In fact, emphasis is laid on datasets which are large in dimensions. This science will also be making use of the intensive computational methods. You can also make use of the classical methods.

These classical statistics models are based on the differential equations or also on the hierarchical Bayesian models. You will notice that there are many overlaps existing in this field. The reason behind this is that statistics computing is created on traditional areas. These traditional areas include- linear models, survey sampling, mathematics statistics, time series and many more. This field of statistics also lays emphasis on many diverse fields which includes intrusion detection, bioinformatics, finance and climatology. You will find number of books on this topic online. You can also log on to the amazon.com and search through keywords. There are many authors which have written on this field of statistics.