Training Courses Public Training -- Scheduled Finance with R Location: London, UK Dates: 2010 November 15-16 A two-day training class where Patrick Burns will be a guest lecturer. For details see: http://www.optirisk-systems.com/events/rtraining.asp Public Training -- Proposed Statistical Programming in Finance with R London, UK, over two days -- dates to be determined. Contents: The course focuses on programming in R. The primary aim is: • For participants to be competent and confident in programming with R by the end of the course. A large portion of the time will be doing exercises that pertain to R concepts. The exercises are generally related either to resampling (statistical bootstrap, random permutation test, cross validation), or stochastic optimization techniques (genetic algorithms, simulated annealing). Hence the exercises perform the double duty of teaching both R and a resampling or optimization technique. Most datasets will be from the field of finance. The additional aims of the course are: • To instill an appreciation for the value of resampling methods, and to encourage their use. • To highlight when stochastic optimization methods are of use, and to show the ease with which they can be implemented in R. Preparation: To take the course you need to bring your own laptop with R installed on it. While no knowledge of R is absolutely required before the course, at least some familiarity is a good idea. A Guide for the Unwilling S User is a very brief introduction to the basic concepts. Some -- but certainly not all -- of The R Inferno will be covered in the course. An introduction to resampling (in R) is The Statistical Bootstrap and Other Resampling Methods. An introduction to stochastic optimization (without R) is An Introduction to Genetic Algorithms. Reading all of these documents should be more than sufficient to prepare you for the course. Bespoke Training Burns Statistics provides in-house training classes tailored to your needs. You can arrange for an entirely bespoke topic, or choose from the selection below. Statistical Concepts Quantitative Tools for Market Prediction Programming in R and S-PLUS Optimization Using Random Algorithms Statistical Concepts Typically 1 or 2 days in length. Statistics is perhaps the most popular topic to dread. It is, however, extremely useful. This course focuses on the concepts of statistics using concrete examples -- mathematics and formulas are pushed as far into the background as possible. Who should attend? • Managers • Traders • Salespeople • Financial professionals • Anyone who encounters randomness The objectives of the class are to: • Become more comfortable with statistics • Be able to understand simple statistical arguments • Learn some common statistical jargon • Be able to recognize the role of randomness A sample outline: • What is statistics? • Why do people find statistics hard? • Examples of randomness • How do statisticians think? • How has computing changed statistics? • Common statistical procedures Quantitative Tools for Market Prediction Typically 1 day in length. While not revealing a few proprietary techniques, this class covers problems, methods and testing when developing quantitative models for market prediction. This provides a unique perspective based on several years of practice, and a strong, pertinent academic background. The objectives of the class are to: • Understand the problem of market prediction • Generate prediction ideas to try • Learn how to evaluate strategies A sample outline: • The Efficient Market Hypothesis • Sources of inefficiency • The computing environment • Function optimization: fitting the models • Classification of prediction models • Survey of prediction models • Trouble spots: robustness, overfitting, etc. • Putting it together: portfolio optimization • Evaluating performance Programming in R and S-PLUS Typically 1 to 3 days. Provides a solid understanding of the S language, and how to use it. An emphasis is placed on good programming technique that saves time and complications in the long run. This is led by one of the early developers of S-PLUS who has written in, documented and taught the S language for many years. The objectives of the class are to: • Understand the strengths and weaknesses of the S language • Become fluent in S • Develop good programming habits A sample outline: • Alternatives to S • S basics • Useful S tricks • Designing functions • Debugging • Writing C code for S • Writing documentation • Developing test suites • Version control Optimization Using Random Algorithms Typically 1 day. Surveys the basic types of random algorithms that have been developed, and then goes on to present how you might want to combine them to best effect. The class is based on more than a decade of developing random algorithms for optimization in a variety of settings. The objectives of the class are to: • Learn when random algorithms should be used • Discover the types of random algorithms that exist • Match the algorithm to the problem at hand A sample outline: • Alternatives to random algorithms • Constraints • Reasons for and against random algorithms • Survey of the types of random algorithm • Advantages and disadvantages of the algorithms • Simulated annealing • The standard genetic algorithm • An improved genetic algorithm • Specializing the algorithm to the problem If you would like to be informed of news from Burns Statistics, please register your name and email address. You can send comments to patrick@burns-stat.com.