Transcript Slide 1
The Art in the Science of Statistics at Macquarie University Protein Structure Prediction from amino acid string to three dimensional structure MVLSEGEWQL Databases VLHVWAKVEA DVAGHGQDIL … AKYKELCYOG + Algorithms Software = Proteins Developing a sequential protein structure prediction algorithm SAINT Mathematical Finance and Statistical Modelling • 1634: Tulipomania in Netherlands • 1636: The Amsterdam stock exchange issued option contracts allowing tulip bulbs to be purchased for a fraction of the price of a real tulip bulb. 1 tulip bulb = A Farm, Live stock, or Life??? If a future market price of a tulip at option maturity time T, is going to be greater than some value K specified in the contract, then the buyer of the contract gets a payoff ‘Market Price of a bulb at T’ –K , otherwise the premium paid for the contract is lost. These options are called Vanilla type options ( or, European Options). Determining Aircraft movement using sound Aircraft Sonogram Listen Compare MI-EL (developed at MU) with other popular Reconstruction Algorithms in Emission Tomography iter 8 iter 16 iter 32 iter 64 2500 MI-EL 2000 MEM 1500 1000 SPS 500 ICA 0 smoothing parameter h = 10-5 The effect of driver’s age on occurrence and size of motor vehicle insurance claims Demand for electricity varies by Month, Type and Time of Day January February March April May June 3000 2000 MWh 1000 0 -1000 -2000 1 7 14 22 30 38 46 1 7 14 July 22 30 38 46 1 7 August 14 22 30 38 46 1 7 September 14 22 30 38 46 1 7 October 14 22 30 38 46 1 7 November 14 22 30 38 46 38 46 December Mon Tue Wen Thu Fri Sat Sun Pub 3000 2000 MWh 1000 0 -1000 -2000 1 7 14 22 30 38 46 1 7 14 22 30 38 46 1 7 14 22 30 38 46 1 7 14 22 30 38 46 1 7 14 22 30 38 46 1 7 14 22 30 Local Correlation plots This graph displays Seasonality Stationarity Autocorrelation Interpretation. Stationarity: Splots have similar shapes. Seasonality: Repeating patterns at regular lag intervals. Autocorrelation: High autocorrelation is indicated by brown/red colour. No autocorrelation is indicated by dark blue colour. High autocorrelation is present at regular intervals along both axes. High autocorrelation at regular lags indicates seasonality. High autocorrelation at regular intervals along CO2 axis fluctuations within seasons. Environmental Statistics An objective is to investigate the relationship between phytoplankton and their environmental habitat using multivariate statistical methods. 2 1 0 -1 -2 -3 2000 2001 2002 2003 2004 2005 2006 2 1 0 -1 -2 -3 2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2 1 0 -1 • The first three principal components explain the overall abundance of phytoplankton studied and the seasonal patterns of a particular genus or an assemblage of various genera that prefer to live in a similar environmental habitat, either in warm or cool water. • The study of seven-year temporal patterns of phytoplankton reveals general patterns of phytoplankton in the reservoir. Such patterns respond to a certain range of environmental factors. This range may be used as indication of a change in phytoplankton composition and abundance. Thus, the environmental factors associated with a situation where one organism invades original inhabitants may be identified. -2 -3 Sangdao Wongsai, PhD student, [email protected] Blue-green algae Grapharti Screen shot of Grapharti from ninemsn daily news polls. Variation in temperature Birthweight versus Gestational age by Birth outcome Birthweight versus Gestational age by Method of delivery stillbirth 6000 6000 normal normal caesar morbid bwt birthweight (g) 5000 5000 other death method of birth outcome delivery 4000 4000 3000 3000 2000 2000 1000 1000 0 0 20 20 30 30 age (weeks) gestational gage 40 40 Seeing Things? The artists Guess who!