196 reputation
5
bio website
location
age
visits member for 5 months
seen Mar 28 at 1:43

Numerical methods for function approximation, predictive analytics, multivariate statistics, data mining


Mar
28
answered What types of neural networks are most appropriate for trading?
Mar
28
answered What is the best way to “fix” a covariance matrix that is not positive semi-definite?
Mar
28
revised Cleansing covariance matrices via Random matrix theory
models
Mar
28
answered Random matrix theory (RMT) in finance
Mar
28
revised Cleansing covariance matrices via Random matrix theory
grammar, clarity
Mar
28
answered Cleansing covariance matrices via Random matrix theory
Feb
22
answered How to cluster ETFs to reduce cardinality for portfolio selection
Jan
29
comment Inferring signals in absence of sign of principal components (PCA)?
The sign of the same loadings can be different across software packages. One thing I have observed across packages is that if the sign of the largest abs(loadings) on the first PC in the range ~0.8-0.9 are negative, then these would be positive when determined using another program.
Jan
9
revised Optimizing Principal Component factor weightings over time
added 77 characters in body
Jan
9
answered Optimizing Principal Component factor weightings over time
Jan
2
awarded  Revival
Dec
31
answered How to normalize technical indicators for machine learning?
Dec
31
comment Machine Learning on matlab 2010
Here's the recent 2012 neuro-wavelet paper on stock time series data I was referring to.
Dec
28
comment How to quickly estimate a lower bound on correlation for a large number of stocks?
If you expand the above, would it be possible to derive an analytic solution for simulating correlated data based on a correlation matrix (any size) with arbitrary correlation values? That is, for example, start with a $30 \times 30$ $\mathbf{R}$ matrix with balanced coefficients. Then apply Cholesky factorization and /or Iman and Conover's approach for simulation. I think what always happens with arbitrary corr values is that $\mathbf{R}$ is not positive definite. Given this, what rules are there for properties of $\mathbf{R}$ when simulating correlated data?
Dec
28
answered What are the general limitations of Gaussian copulas with regards to the range of joint pdf's it can approximate?
Dec
28
comment how to choose top n assets?
Wouldn't it be better to invest in the GMV portfolio which is not based on expected returns - since returns change through time and are unpredictable?
Dec
28
answered Machine Learning on matlab 2010
Dec
9
answered Large (5K-10K) non positive definite (particularly near singular) covariance matrices and treatments for Cholesky decomposition
Dec
8
answered Choosing a weak learner
Dec
8
comment Evaluation volatility with Garch model
There was a post in QF a while ago suggesting at least 5000 data points (in the time series) for a GARCH model