As a neuroscientist, should I even bother learning calculus...

As a neuroscientist, should I even bother learning calculus? Or will all analysis and computation in my field be done by maths / comp people? I function well as a neuroscientist and I have though about taking calc courses, but would it give me anything right away or just the preliminary to study more?

Other urls found in this thread:

nature.com/neuro/journal/v14/n9/full/nn.2886.html
twitter.com/SFWRedditGifs

That's like asking if you should or should not go to the cinema because someone may at one point ask you about your opinion for the movie.
It's a fun thing to do and only takes a few hours off your time.

You probably won't need it, but mitebfun

I'm a neuroscientist also, And I'd recommend being familiar with calc for use in neural oscillator analysis. There are many algorithms that entail area under the curve, for instance, in psychophysics.

I'm a machine learning researching who just casually looks into neuroscience research from time to time, and I definitely have seen plenty of calculus. More than I expected.

Calculus is really a fundamental tool for any kind of scientist, make sure you know it.

What kinds of theoretical stuff do neuroscientists actually work with? I'm really interested and want to know.

I was recently at a topology conference in Austin and spoke with a guy who uses differential geometry to understand changes in the development of brains from children to adults happen.

Neuroscientists confirmed brainlets

Depends what you're doing user. Go read about cable theory

This pic triggers me somewhat

Why?

What do you do as a neuroscientist? Research?

Is it a hard to get into position - where a random individuals out of the top people get into? Can you explain how hard was for you to get where you are? And what other opportunities are there in the field of brain study.

Nope, learning something will actually set you back in your field. Studies show that every time you read a book, you lose $5,000 off of your future salary and 10 points off of your IQ.

I love your post.
How would this type of human be called? It's literally the only thing that makes me laugh.

Someone with good sense of humor

you can't call your a scientist if you don't know calculus

Learn some fucking statistics so that 90% of results in your field won't actually be worthless. You soft sciences people are shitting up the literature because you don't know enough math.

I'm a postdoc researcher. Yes it is hard to get a position like this but it's even harder to get a tenure track prof position and get grants. Knowing people is half the battle so expand your network as best you can. The other half is having unique skills that match up with thhe lab's needa. I expect within the next 10 years there will be plentiful positions in neuro(insert corporate field here). I do data science and marketing consulting as well as my research.

You do not read calculus. Just use Dr. Tai's patented revolutationary method to calculate the area under a curve and you will be fine.

Or better yet, ignore her result and just re-invent it and publish it yourself. No one will notice for decades.

The short answer is that if you want to stick to the basics, then you do not necessarily need to know calculus, or any form of math beyond high school level for that matter. There are plenty of people who deliver perfectly alright work with just conceptual knowledge of what they're doing.

That said, I expect that in the next 10 to 20 years computational neuroscience will take a huge leap forward, and mathematical descriptions of the processes under study will become increasingly important. Such models are already gaining some traction, because they reduce the data to a set of algorithmic descriptors and thereby provide understanding at a mechanistic level. Ultimately, that is a large aim of our collective endeavors: understanding the brain in its entirety. The people who don't know how to properly incorporate or even interpret computational models will simply be left behind.

Aside from that, as our tools to study neural dynamics are becoming increasingly complex, so does our data. A natural consequence is that our analyses are getting more sophisticated, and so to be able to do these analyses you need to understand them and their mathematical basis. Otherwise you run the risk of making improper decisions when it comes to setting the analysis parameters, and as a result may inappropriately interpret the results. A lot of people tend to just use a point and click software package with default settings, but these packages will probably become less widely used when the variety of analyses that is available and their speed of development exceeds the speed at which these packages can be updated.

I've kept this kind of general on purpose because the kind of math that is most heavily relied on is hugely dependent on your field. I can get more specific if you want, but for that I'd need to know what type of stuff you typically (want to) work on.

>90% of results in your field
[citation needed]

And remember, if you post anything that doesn't specifically mention the 90%, it'll be pretty ironic.

If you plan to analyze raw EEG data, you need Calculus and DSP.

no you don't

Okay, if you say so.

You probably won't actually need calculus, but it's very easy and you really should learn it. It blows me away that someone with a science major could somehow graduate without ever taking calculus.

I doubt that calculus will be necessary for your field, but it's a cool area of science regardless and it's very easy

What do neuroscientists do? I'm not quite sure what separates a neuroscientist from a biologist or psychologist and to what degree they're interested in the dynamics of what they study versus simply identifying its makeup/purpose.

If it's categorizing the brain, its structures, their general function, and hoping to connect certain behaviors with certain features of the brain, then you probably don't need it. If it's tirelessly analyzing data, then you only need it if you're creating your own statistical analysis programs/functions.

If, on the other hand, you want to develop a mechanical understanding of the inner workings of the brain, I can assure you that calculus will be incredibly advantageous to building models of different processes/phenomena. Due to its complexity, I don't think neuroscientists focus on this area--maybe it's more suited to condensed matter/statistical physics approaches.

>I don't think neuroscientists focus on this area
We do though. That's the whole point: to not just understand the brain's architecture, but to understand its mechanics.

How on Earth is whether or not to learn basic calculus even a question, then? I don't understand how someone could hope to model something as unfathomably complex as the human brain without a tool as fundamental as calculus.

Because to some degree you can empirically determine the mechanics of a simple system. If I want to know if let's say process X is driving theta oscillations in region Y, I knock out X and see if affects the oscillations in region Y.

>something as unfathomably complex as the human brain
In a lot of cases people study sub-systems in isolation. Understanding the constituent parts of the system is necessary to understand it in its entirety. That doesn't undermine the end goal of a mechanistic understanding of the whole brain. In fact, it's an integral part of it.

I'm not saying that calculus is not necessary for neuroscience as a whole (of course it is), but you can work as a neuroscientist and not use calculus.

No wonder they aren't making any progress if they can get into research without ever knowing Calc1

>aren't making any progress
It's probably the fastest developing field of science out there.

>If I want to know if let's say process X is driving theta oscillations in region Y, I knock out X and see if affects the oscillations in region Y.
You'll need calculus to perform the necessary statistical analyses for that

You need calculus to design the statistical test. To apply the test to your data, you only need to know which buttons to click in the analysis software.

If you don't even understand what you're doing, how are you going to understand the brain
For statistical tests and especially regression analysis, you need to master the underlying theory to interpret them properly

All of that can be done conceptually instead of quantitatively, in principle.

No, absolutely not

That isn't an argument.

Why do you think 70% of psychological research has turned out to be false? Those psychologists could have interpreted their results "conceptually" right?

>It's probably the fastest developing field of science out there.

As a college age student, don't you have to take Calc I for any science degree?

I wouldn't want you as a neuroscientist if you didn't/couldn't complete a class I finished sophomore year of high school.

Ironically, psychologists often get much better education on statistics than do neuroscientists. Papers that criticize the reliability of psychological research tend to come from psychologists themselves rather than from neuroscientists. Psychologists have written similar papers criticizing the stats used in neuroscience, for instance: nature.com/neuro/journal/v14/n9/full/nn.2886.html

Anyway, we're getting side tracked. The only point I wanted to make is that it's perfectly possible to do neuroscience without mastering calculus. Where we do agree, I think, is that if you want to deliver good work, it's probably a good idea to have some clue of what you're doing.

Nah, there are neuroscience masters which you can enter with a BA in psychology. Then you wouldn't touch calc

Is object-oriented programming or data structures & algorithms useful tools in neuroscience? Specifically for analyzing neural time data, such as LFP recordings or Ca2+-imaging?

The brain does difference equations
Its combining two images seamlessly in realtime and using visual cheats somehow
You probably won't be doing any difference/volume calculations in your studies though

>How would this type of human be called?
le sarcastic faggot

Very good post.

Nowadays, more and more computational research in neuroscience seems to be of complex systems and/or systems biology flavor. And, most of the people in the field seem to have backgrounds in math and/or physics.

It won't be necessary when this field is flooded with Pajeets.

This is really surprising. For the longest time I saw neuroscience as a field for the very smartest partaking in the 'softer' sciences

>Is object-oriented programming

OOP is a means to enable dynamic polymorphism and Liskov substitutability. This is very useful for software projects that have to keep track of varying structures that each require potentially arbitrary differences in behavior that won't necessarily be known ahead of time, but it won't matter at all if your project only needs mathematical functions that operate in specified ways on specific kinds of data.

PROTIP:
Don't study calculus if you know you won't be using it, spend the time more wisely.

I would think that it has more to do with the complexity of data that is dealt with in the different fields. In neuroscience programs a lot of time is devoted to DSP and electrophysiology and such, which isn't always necessary for psychologists. In psychology programs there's simply more time to focus on proper stats because of that. In the former field, stats are assumed to be trivial, but that assumption doesn't hold up.

From article
"[...] we reviewed an additional 120 cellular and molecular neuroscience articles published in Nature Neuroscience in 2009 and 2010 (the first five Articles in each issue). We did not find a single study that used the correct statistical procedure to compare effect sizes"

So neuro grads should learn their stats then

Yup. And this point was made by psychologists, no less.

this says it all pretty much

...