In particular, we chose the studies GM05296 and GM13330. The data correspond to two array CGH studies of fibroblast cell strains. We selected a subset of the data set presented in Snijders et al. (2001). If we have time we will also cover this algorithm.) (Aside: DNAcopy implements another useful signal processing algorithm, called circular binary segmentation. We load an arra圜GH dataset from a Bioconductor package, DNAcopy. Glossing over the technological details, if the signal is around 0 it means the sample has the same number of copies of DNA as the reference consistent stretches of higher or lower values than 0 indicate amplifications and deletions, respectively. Array comparative genomic hybridization (arra圜GH) is a microarray technology which compares, at a set of regions tiling the genome, the amount of DNA from a sample compared to a reference. Here we will present the Hidden Markov Model (HMM), and apply it to detect copy number variants in arra圜GH data. Blei and Ghahramani maintain C, Matlab codes on their website as well.Hidden Markov Models and Dynamic Programming Michael Love There is a R package called lda and topicmodels for most of the "good" models.
A few impressive ones are: The infinite hidden markov models, Time sensitive Dirichlet Process Mixture Models. You can find here the specific HMM models that you require. Time series and sequential data models (HMM specifically)Īpart from Jordan and Blei, the other prolific research is Zoubin Ghahramani (and his coauthor Beal). Its hard to find specific references there as he publishes so much! Another great reference in the same field is his advisor, Michael Jordan's, website. He is doing some great work which is very general, so it might not be surprising if he has already done something in finance. The specific references, the slides for the presentation, and more complicated models can be accessed from his website.
You can watch a nice presentation by David Blei (great presenter, apart from his awesome!! research) here. They have shown to be state-of-the-art in topic modeling. They use the latent variable formulation, which is very similar to that of HMM. These approaches are not used in finance (I am not aware, as I don't work in specifically in finance), but they have very general applicability. Dynamic topic models (IMHO, most suited for your case, without much domain knowledge)Ĭ. Topic Models (used to find patters in a set of documents and/or information retrieval)ī. I think a few methods that can be used, but not designed specifically for you, are as follows:
0 kommentar(er)
