Tag Info

New answers tagged

0

To close this question. Steps used, in short : get matrix N x M where N - number of assets, M - number of history samples normalize all samples using logarithms and mean to have returns instead of some asset specific values obtain covariance matrix, or correlation, if you want to avoid influence of volatility solve eigenproblem using SVD and Jacobi's ...


1

Q #1: I'm not sure if you have the answer quite right. The signs for the loadings are arbitrary, but you cannot take the absolute value. You can multiply by -1. Q #2: It might be helpful to think about what PCA is actually doing. This paper might be helpful: http://arxiv.org/pdf/1404.1100v1.pdf (A Tutorial on Principal Component Analysis by Jonathon ...


0

A # 1 : Several replies from the following topic answer my Q # 1 - yes, if I take only one dimension after PCA then I can simply make all vectors positive, hence take only absolute values by module. If I take several dimensions then entire vector needs to be reverted and each value inside particular vector can be multiplied by -1 because reversion of ...



Top 50 recent answers are included