<t>What you are mentioning has been quite a hot topic in Machine Learning community, the term used for this is 'Online Learning'. There are online implementations for many machine learning algorithms, and svm is no exception. Once you dive into this however, you need to be really careful about over-...

<t>There are many approaches to this problem, and none of them can be successful without taking into account the competition. If there is no competition for the opportunity you are going after, then you can assume 100% fill rate. It gets interesting when you are going after the same opportunities wi...

It's probably OTC trades. Some exchanges register these trades on the wire, and this might result in trade prices within bid-ask spread

<t>Computer Vision..maybe you can use the kalman filter implementation, and modify some of the tracking methods. if you force it, maybe segmentation routines can be used, thinking panel data as an image? but I don't see the point of using a computer vision library if you only need the machine learni...

<t>I think the content of the spam has immensurable claims, like 'finding the result amazing'.. It would have worked better if it had more direct claims like 'landed a job in XYZ Hedge Fund/Bank/Firm as the Head of ABC' ...it might have a better impact in case the fake claim is a bigger one, I perso...

no link for me?or for noone?

haha, from receptionist to hedge fund professional.. either it's that good, or you become a professional of a different kind

<t>Hi Brian,Your question was clear actually, but the answer depends on your assumptions. The correlation and covariance are both related statistical measures that summarizes your data. You can see correlation as the adjusted(normalized) covariance. I think the figure below explains what I was tryin...

<r>The "proper equation" will be highly biased toward your view of the market, which is your model, and which is always wrong.The linear relations that you have identified in the last two years will probably not be valid that day (like one of the stocks going wild etc)..It makes sense to start with ...

<t>I have to admit that it is quite a harsh reply to someone newly starting the field. I don't mean to offend you, but I don't agree with many points in your post. Being in the machine learning and pattern recognition field for some years, let me try to demystify some of the concepts in your reply, ...

<t>Hi Martin,I am doing a PhD. on Machine Learning Techniques in Finance, in Intelligent Systems Lab, UVA, Amsterdam. I have been struggling to find a mid-point between finance and Machine learning as well. I have to admit it's not an easy challenge to satisfy both fields.The subject specialization ...

Thanks for the pointer to the paper, I have to admit that I haven't seen a formal comparison between Cuda and OpenCL until now. My impression was mostly formed by reading through the forums

<r>If highest performance, including latency is an issue for you, I recommend staying away from the OpenCL. Cuda has a lot better overall performance, at least for now.That being said, you can check the official book from Nvidia, which is accompanied by many good examples:<URL url="http://developer....