If you need some linking between machine learning and speech recognition, there is a theoritical and state-of-the-art paper describing speech recognition (mainly acoustic models) techniques in the language of machine learning researchers. http://research.microsoft.com/apps/pubs/default.aspx?id=189008 This is far from implementation level though...

Human RBCs or from other vertebrates? For all mammalian RBC just use a hypotonic solution and all Hb will be released. This is a classic method to obtain the cytoskeleton and membrane proteins.

You have considered some very interesting points here Jismine in which I see different frames of reference being used by the supervisors of a PhD. As a piece of writing the PhD should be assessed not only for the discourse in answering the thetic question posed, complete with bibliography - but also for its standard of precise writing, style, tone, sense of audience, rigorous and unambiguous argument with examples, logic and flow - and other basic necessities such as grammar and syntax. I would also allow some lenience in the use of "poor writing" for the initial draft. The student will usually have the thesis professionally scrutinised by a proof-reader, editor (usually the same person) so as to present a quality paper expected of a learned person, before final submission of the paper. My writing colleagues will always make use of a "critical friend" to read through a text and advise where the text fails, in any respect, in meeting its intention and targeted audience or by suggesting new material to be included which may support the text. This applies to all genres. As a PhD is an analytical work of research and argument against givens, to prove a new idea, then both argument and language must be used so as to avoid misunderstanding of the text by the intended reader. The bibliography is a set of further reading which may be used by the reader to site the writer's text. In "everyday" speech and writing, the 'continuants' are acceptable only when they are not considered important by the receiver. The receiver will seek further clarification of the meaning and intention of the communication as required. Assumptions can mean virtually anything and may cause severe consequences. I liken this to your earlier reference to 'reader responsibility'. And so on . . . ;)

Boyan, thank you for this comment. Do you know how extensive the ion exchange is, and is this in general for all ILs? I am asking because I adapted my purification procedure from a few different literature procedures, including the addition of drying agent. I questioned ion exchange at first, but the paper reported drying without ion exchange. We have not seen different behavior of our organic compounds in ILs with sodium sulfate added compared to those purified without its addition, (except water content lower) but I am curious to know more. The sodium sulfate is only present in an IL-CH2Cl2 solution for less than 20 minutes in our procedure.

@Purva... Yes, SPIONs will finally yield Fe3+ or Fe2+ ions.. but in your case this process is delayed, due to the presence of surfactant or protein protective layer...

Assuming you use the same burette, reagents etc and you only change the reading you report the measurement error should be the same. Need also to report the error in liters.

:(

TEM is best way to measure the particle size. When you do the synthesis of nanoprticle it do happen that you will get multi dispersed particles. Just try to synthesis the another new particles or you can take the particle size which is more common .

Hi Wael, I haven't made an SSR genotyping so I am not sure what the protocol for your samples are but, the times I have run gels and gotten some weird migration as your gel show have been because my samples were too salty. One thing that might help is to leave an empty well in between samples. The reason to leave the empty well is to leave space for the salt to diffuse. I hope it helps!

A long time ago I posted some chaotic time series on the web: http://www.physics.emory.edu/~weeks/research/tseries1.html one of which I believe is low-dimensional chaotic data, from an experiment. (In fact several of the time series are experimental, but I think only one of them is chaotic.) I never did demonstrate completely that it was chaotic, it was only a side project when I was a graduate student, and now I don't do time series analysis. But the data might be of use to you, you can download it from the link.

Ahmed. Well, I didn't mean to suggest that the Copenhagen or Many Worlds Interpretation is wrong because it seems ridiculous anymore than I would say that relativity is wrong because some aspects of it seem counter-intuitive. My point was that some physicists find the MWI absurd and others find the Copenhagen approach ridiculous, which only confirms my appraisal that, at the level of electrons, there is no consensus as to whether one can say that the wave has a (definite) structure in some external or ontological sense or meaningful sense. What the wave is doing and why it is doing it or why it takes on properties of a more tangible particle when observed seems to be a mystery. Sure, we can make measurements of a table just as we can piece together relative frameworks, and make sketches of both. But again, Einstein would say that “As far as the laws of mathematics refer to reality, they are not certain.” It is one thing to state that at a macro level there appears to be consistent structure, and quite another to speak in terms of absolutely and infinitely precise measurements, and quite another to suggest that there is some ultimate, non-local, stationary frame of reference in which to measure all movement in the Universe. I am not sure what is meant exactly when people speak of absolute truth . And if there is such a thing, is there supposed to be some sort of esoteric link between it and absolute moral “truth?”

I can not tell about algorithms, but here is a tool called HIVE, http://hive.apache.org/ for big queries on big data. Facebook uses it for analyzing user generated data, but it runs over Hadoop, so the warning of prof. Steffenel is relevant. If you look more about Hive, may be you can find the algorithms used there.

For python, I will not restate some of the excellent resources above, so I will list one that I have had good luck with: http://learnpythonthehardway.org/ You can pay for the videos, but just progressing through the problem sets in order is very beneficial to learning.

Ramana, Mantoux test is unreliable in detecting latent tuberculosis infection (LTBI) in countries where all children are vaccinated with BCG. This is because the Mantoux test positive in everyone who was exposed to the vaccine. However, with the advent of IGRA, two assays, T-spot and QFT-TB Gold, the number of false positives have significantly declined. This is a relief for countries which administer BCG to children at birth. The only downside of IGRA is cost. Thanks, Ben.

Some details taht can be useful to the readers of an survey paper: 1- Cite together papers dealing with the same method of discovering association rules. Some present the first draft of the methods, another the mature method, others new applications of the same methods. It is easy, generally those papers share at least one author. 2- Provide one or more taxonomies of methods for association rule mining. Each taxonomy should be based in some important criteria (for example, classification of methods by algorithm, or by knowledge representation, or by problem niche addressed). May be there is a taxonomy already created, then you should group papers into classes; may be there is no taxonomy, and you should propose one and then you shoul group papers into classes. It is easier if you did step 1. 3- Discuss advantages and shortcomings of each method. I suggest to review ACM Computing Surveys, there you can find many survey papers: http://dl.acm.org/citation.cfm?id=J204&picked=prox I recommend you to follow the example of this fine survey paper: A survey of online failure prediction methods, http://informatik.hu-berlin.de/Members/salfner/publications/salfner10survey.pdf

As a biochemist/molecular biologist, I have found the learning curve to be much easier with python. Many of the previous entries list the various reasons for this. In particular, I am very appreciative of the helpful community at large, many of which do not participate in the academic/science realm, but are willing to assist me in solving large data generated problems that seem to be the trend with my own datasets. That being said, I started with perl and still utilize simple one-liner scripts to perform specific tasks in my workflow.

Thanks, Santosh Maurya.

I have three duplicate answers because my computer did not appear to register my answer.

They could tell me a simple protocol of wester-blot from kidney extract in non-reducing conditions. Regards