I'm trying to make a formula!

Hi, I have always been curious on how scientists like Albert Einstein create formulas and equations that solve problems and support their theories. So I wanted to give it a try myself. I created a sample of my formula which I know it is 100% wrong but, I wanted to present it anyways so I can give an example of what I want to achieve.

PROBLEM: Babies always need something, they cry to get our attention and expect us to give them what they need on the spot. The problem is that we don't always know what they need because they are not able to talk in our language yet. So I want to create a formula that will help us decide what the baby needs through a set of variables and finding the percent chance of each so people can do a process of elimination based on these chances.

Overall, I want my formula to be used so people can check for the obvious reasons(variables with high percentages) and work their way down to quickly find the reason why their baby may be crying.


(variables)

hungry =H
sleep = SL
attention = A
dipper = D
shower = SH
pain = P

(/variables)


(example)

H + SL + A + D + SH + P
------------------------ x 100%
24

(/example)

HOW IT WORKS: Users will plug in the amount of time(hours, minutes, etc) for each variable based on how much time has passed since the activity has been done. Then divide each variable by 24 hours and then times the value to 100%.

NOTE: if you choose minutes you have to do that for all variables and etc for other forms of time.

(example2)

3H + 0SL + 2A + 3D + 0SH + 1P
------------------------------- x 100%
24(hours)

RESULTS:

H = 12%
SL = 0%
A = 8%
D = 12%
SH = 0%
P = 4%


DESIGN: I want my formula to be flexible, in which the user can exclude variables that hold the value of 0 so they can shorten the time it takes to solve the formula(like in some math equations). Also, I want users to be able to create and plug in their own equations because these variables are limited and sometimes the user's needs can be different from others.

(/example2)

So I know this is completely wrong but that's what I am trying to do. Create a formula that will make the process of finding what the baby needs in a more easier way by inputting the amount of time that has passed(hours, minutes, etc) before the given variable has been checked or done for the baby. This will then output the percent chance of each variable so that the obvious action that your baby needs is clear.

I will appreciate any help that is given

Thank you,

Mendez.Angel

Formula Design generally takes a lot of trial and error. Please be careful when experimenting on peoples' children.

Generally, you want to make sure your theory is sound BEFORE you start writing things down. Once you have a cogent theory, try to build a model and do your best to DISprove it.

However, and somewhat to the consternation of his peers, Henry Eyring, one of the premier chemists of the 20th century, occasionally worked the other way around, i.e. See what works and then invent a cogent theory.

Social sciences (like tending babies) can be very much trickier than fields were the experimental objects behave more reliably.

The model does not make sense from an algebraic standpoint, how does (3H+1P)/24*100 give any values whatsoever for H and P?

I think what you're trying to say (at least from example 2) is that the probability a baby needs care in the form of x is T(x)/24, where T(x) is the number of hours since care of the form x has been given.

So many problems with this, the greatest of which is the assumption that all types of care are equally needed. Ex: Pain and Food...

Second, this model will say that if a baby is crying and hasn't had pain for 24 hours, then it is certain to be the cause!

daon2 said:

The model does not make sense from an algebraic standpoint, how does (3H+1P)/24*100 give any values whatsoever for H and P?

I think what you're trying to say (at least from example 2) is that the probability a baby needs care in the form of x is T(x)/24, where T(x) is the number of hours since care of the form x has been given.

So many problems with this, the greatest of which is the assumption that all types of care are equally needed. Ex: Pain and Food...

Second, this model will say that if a baby is crying and hasn't had pain for 24 hours, then it is certain to be the cause!

Click to expand...

"I think what you're trying to say (at least from example 2) is that the probability a baby needs care in the form of x is T(x)/24, where T(x) is the number of hours since care of the form x has been given." - daon2;320171

Yeah that's something that I was looking for. Now I know it is completely wrong because of different factors come into play for example, a baby may eat 6 times a day or sleep 3 times a day for a certain amount of hours and so forth. If you needed to make a formula for this, how would you go about in implementing these factors and creating a "test formula" so it can be tested?

"Second, this model will say that if a baby is crying and hasn't had pain for 24 hours, then it is certain to be the cause!" - daon2

The reason why I added pain to my equation is not because I am assuming the baby had pain or an injury for the past 24 hours. It is to check if the baby does have an injury or is experiencing pain of some sort. That is sometimes the case when a baby cries, they may have pain(injury, illness, etc.) that you have not been aware of. So the equation is to let the users(or parents) know that maybe they should check for these things.

Thanks for the reply,

Angel Mendez

tkhunny said:

Formula Design generally takes a lot of trial and error. Please be careful when experimenting on peoples' children.

Generally, you want to make sure your theory is sound BEFORE you start writing things down. Once you have a cogent theory, try to build a model and do your best to DISprove it.

However, and somewhat to the consternation of his peers, Henry Eyring, one of the premier chemists of the 20th century, occasionally worked the other way around, i.e. See what works and then invent a cogent theory.

Social sciences (like tending babies) can be very much trickier than fields were the experimental objects behave more reliably.

Click to expand...

Thanks for the reply,

"Social sciences (like tending babies) can be very much trickier than fields were the experimental objects behave more reliably." - tkhunny

yeah that is true but, that's why I wanted to come up with values that are percentages. My goal is not to be 100% accurate, I just want to display the likely hood for each baby behavior(most common ones or the ones that the users plug in).

PS: don't worry, I'm not going to starve or harm anyone's baby(s)

Mendez.Angel said:

Hi, I have always been curious on how scientists like Albert Einstein create formulas and equations that solve problems and support their theories.

Click to expand...

In addition to a university degree in your field of study, it takes years of learning mathematics. After that, stuff begins to happen naturally because scientists like Albert Einstein are doggedly stubborn and refuse to give up. :cool:

Mendez.Angel said:

Hi, I have always been curious on how scientists like Albert Einstein create formulas and equations that solve problems and support their theories. Mendez.Angel

Click to expand...

Not all problems are amenable to arithmetic formulas. There are other ways to find answers that are either correct or probably correct. Differential diagnosis is one. That is not a numeric formula, but it is mathematical (in principle) because it describes a logical process to eliminate possible alternatives. The problem you propose seems ideally suited to differential diagnosis.

The idea is that there are a relatively number of reasons that the baby is crying. One is that the baby is sleepy. If, however, the child is six months old and has been awake for fewer than 10 minutes, I think we can rule that possibility out. Another is that the baby is hungry. If it ate less than half an hour ago, that possibility also seems unlikely. Try burping the baby. The process is not explicitly quantitative, but effectively is a qualitative Bayesian process that reduces certain a priori probabilities to close to 0.

OK, I get it. So I can't really make a formula for this, so I'm pretty much done. Thank you all for your time, I know that not a lot of people would have taken the time to answer this.


Thanks again,

Angel Mendez