# Is my modeling approach correct?

• SMA_01
In summary: Yes, you would need to define new variables xp and yp, representing the desired population levels for blue whales and fin whales, respectively.You would then incorporate them into the equation by using them as the target population levels for harvesting, as explained in my previous response. This will help you determine the optimal levels of blue whale and fin whale populations that will result in the maximum revenue.
SMA_01
Ecologists use the following model to represent the growth rate process of two competing species, x and y:

dx/dt=r1x(1-x/K1)-α1xy

dy/dt=r2y(1-y/K2)-α2xy

Let
x= number of blue whales (measured in 100,000's)
y=number of fin whales (measured in 100,00's)
t=time measured in centuries
r=revenue obtained from harvesting (measured in 1,000,000 of dollars per year)

A blue whale carcass is worth $12,000 and a fin whale carcass is worth$6,000. Assuming that controlled harvesting can be used to maintain x and y at any desired level, what population levels will produce the maximum revenue? (Once population reaches the desired levels, the population levels will be kept constant by harvesting at a rate equal to the growth rate).

The equation without harvesting are
dx/dt=5x(1-x/1.5)-xy/10

dy/dt=8y(1-y/4)-xy/10
_________________________________________________________
I am a little unsure about my equation for revenue- more specifically, I don't know how to incorporate time.

I have r= 12,000x + 6,000y, but that doesn't seem right because I neglected t. So I thought maybe this is more accurate:

r=12,000 (x/100t) + 6,000 (y/100t) (since t is measured in centuries)

Is this correct? Or am I completely missing the point?

If the population is maintained - why would the harvest rate depend on time?

The variables 'x' and 'y' represent the total population of blue whales and fin whales, respectively.

You cannot use 'x' and 'y' in your revenue equations without some modification, since you are not harvesting the entire whale population at any given time.

What you are trying to do is manage the two whale populations so that, with harvesting, the number of whales in each population will not grow beyond certain figures, call them 'xp' and 'yp'.

So, in a given year, if the population of the blue whales, x(t), exceeded the desired population, xp, then the excess would be harvested, yielding a revenue of [x(t) - xp]*12000/1000000 millions of dollars. The blue whale population would be reset to x(t+1/100) = xp for the following year. A similar analysis can be done for the fin whale population.

Now, the trick is to select the values of xp and yp which will maximize the amount of revenue obtained from the harvests. According to the OP, the yearly rate of harvest should be equal to the growth rate of the population. For the blue whales, this harvest rate would be [5xp(1-xp/1.5)-xp*yp/10]/100 blue whales per year.

Wouldn't it be rather more simple and easier to get the big picture to split up the actual growth rate into the expression for the natural growth rate and the harvesting rate?
That is, writing for dx/dt:
dx/dt=5x(1-x/1.5)-a_1xy-r_1*x
Obviously, you can gain an equivalent formulation in your terms, but why do you want the harvesting rate to appear as a variable in the denominator in the expression in the parenthesis.

I believe my approach will make the algebra simpler.

Agreeing with arildno. These are equations close to though not quite the same as classical Volterra-Lotka equation. They probably predict, like those, cycles around a fixed point. However the problem essentially is about the fixed point. Or rather the new fixed point when the animals are harvested.

Firstly how do you find the fixed point? (There is a trivial one at 0,0 for the undisturbed system.)

You don't need to find it for the undisturbed system, though interesting, only for the system where there is harvesting.

So you have to add to your equations a term for the effect of harvesting.

Arildno suggests one and it certainly simplifies the algebra.

However you have to decide what is your model for the harvesting! Arildno's model is like when the animals are hard to catch and the yield is proportional to the effort and you catch all the animals you can so there is a term proportional to x. You'd have to make more complicated assumptions for 2 species. I'd think a better model is when you can just decide how many animals you catch. In that case the last term in each equation is a constant (two different).

So you'd get the equilibrium populations of x and y as function of the two harvesting rates. I guess there is always only one relevant solution but that should emerge from the math. Then you'll have to determine what among the possible populations is the best \$ yield.

Interesting question, keep us posted when you have progressed.

Modelling harvesting as a constant makes the actual harvest insensitive to population fluctuations.
Rather than the "hard to catch" idea, a term rx is the simplest way to model a case where you instantaneously adjust harvesting to balance the natural growth.

SteamKing said:
The variables 'x' and 'y' represent the total population of blue whales and fin whales, respectively.

You cannot use 'x' and 'y' in your revenue equations without some modification, since you are not harvesting the entire whale population at any given time.

What you are trying to do is manage the two whale populations so that, with harvesting, the number of whales in each population will not grow beyond certain figures, call them 'xp' and 'yp'.

So, in a given year, if the population of the blue whales, x(t), exceeded the desired population, xp, then the excess would be harvested, yielding a revenue of [x(t) - xp]*12000/1000000 millions of dollars. The blue whale population would be reset to x(t+1/100) = xp for the following year. A similar analysis can be done for the fin whale population.

Now, the trick is to select the values of xp and yp which will maximize the amount of revenue obtained from the harvests. According to the OP, the yearly rate of harvest should be equal to the growth rate of the population. For the blue whales, this harvest rate would be [5xp(1-xp/1.5)-xp*yp/10]/100 blue whales per year.

Thank you. So would I need to define two new variables xp and yp?

arildno said:
Modelling harvesting as a constant makes the actual harvest insensitive to population fluctuations.
Rather than the "hard to catch" idea, a term rx is the simplest way to model a case where you instantaneously adjust harvesting to balance the natural growth.

I was going to argue about the considerations that introduces, but I have re-read the question and it says "Once population reaches the desired levels*, the population levels will be kept constant by harvesting at a rate equal to the growth rate". That is, the added term in each equation is a new constant (different for each) multiplied by dx/dt or dy/dt. The result is that the each RHS is the same for with as without harvesting except it is multipled by a constant when there is harvesting, so when you have solved for without harvesting you easily have the solutions for with.

A slightly impractical model IMO but we are talking about principles.

(*? However it does not do that by itself, so the question can only mean it is brought there by some unspecified means that is outside the question.)

SMA_01 said:
Thank you. So would I need to define two new variables xp and yp?
see ^^

Last edited:

## 1. What is the best way to evaluate the accuracy of my model?

The best way to evaluate the accuracy of your model is by using a combination of metrics such as mean absolute error, root mean squared error, and R-squared. These metrics will give you a comprehensive understanding of how well your model is performing.

## 2. How do I know if my model is overfitting or underfitting the data?

Overfitting occurs when the model performs well on the training data but fails to generalize to new data. Underfitting occurs when the model is too simple and does not capture the complexity of the data. You can determine if your model is overfitting or underfitting by using techniques such as cross-validation and learning curves.

## 3. Is it necessary to use feature selection techniques in my modeling approach?

Feature selection is an essential step in building a model as it helps to eliminate irrelevant or redundant features, which can improve the model's performance. However, it is not always necessary, and it depends on the complexity of the data and the type of algorithm being used.

## 4. What are some common pitfalls to avoid in modeling?

Some common pitfalls to avoid in modeling include using a biased dataset, not properly preprocessing the data, and not considering the assumptions of the chosen algorithm. It is essential to thoroughly understand the data and choose appropriate techniques to avoid these pitfalls.

## 5. How do I choose the best algorithm for my data?

Choosing the best algorithm for your data depends on several factors, such as the type of data, the problem you are trying to solve, and the performance metrics you are trying to optimize. It is recommended to try multiple algorithms and compare their performance to determine the best one for your specific data and problem.

### Similar threads

• Calculus and Beyond Homework Help
Replies
7
Views
2K
• Calculus and Beyond Homework Help
Replies
1
Views
2K
• Calculus and Beyond Homework Help
Replies
1
Views
7K
• Calculus and Beyond Homework Help
Replies
1
Views
2K
• Calculus and Beyond Homework Help
Replies
7
Views
5K
• Differential Equations
Replies
1
Views
6K
• Differential Equations
Replies
2
Views
2K
• Math Proof Training and Practice
Replies
46
Views
10K
• MATLAB, Maple, Mathematica, LaTeX
Replies
10
Views
2K
• MATLAB, Maple, Mathematica, LaTeX
Replies
0
Views
2K