Carlos Valdez

Dear Math, please grow up and solve your own problems. I'm tired of solving them for you.

[Carlos Valdez @c_eleazarV]

@Ash_Kalra @Wimbledon We're all watching it at work!

[Carlos Valdez @c_eleazarV]

@holston_sebaugh It won't open for me either. Maybe you can show it to me later.

[Carlos Valdez @c_eleazarV]

@austinharper59 This is a pretty good intro to ode in MATLAB. It's always good to compare all these different explanations and find an approach that works for you.

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB You're right, there are a lot of examples. They're more complex than what we're doing but it's still a good reference for understanding how different DEs work in MATLAB.

[Carlos Valdez @c_eleazarV]

Here's an example of how DEs can be applied to modeling data https://t.co/N4aK7U0mG2 #math3185

[Carlos Valdez @c_eleazarV]

@austinharper59 Good recap. This reminds me of the methods we would use to find the area under a curve before we used integrals. As we set dx to be smaller, we would obtain a better approximation.

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB This is a good explanation. To be honest I haven't had that much time to actually fully understand the process. It now makes more sense about why we choose the most negative number as our pivot column.

[Carlos Valdez @c_eleazarV]

@holston_sebaugh Nice video! I can see why we do the geometric method to get a nice visual understanding of linear programming. If we just jumped straight into the simplex method it would probably be a bit confusing.

[Carlos Valdez @c_eleazarV]

This article talks about gradient descent algorithms. It's very detailed, but it's cool to skim through and see the different algorithms that use this concept.
https://t.co/SYydX51RDp #math3185

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB It's always nice to be able to visualize concepts like this. It's like rolling a ball on an incline surface and watching where it eventually stops. The ball does not do anything if you drop it on a flat surface.

[Carlos Valdez @c_eleazarV]

@holston_sebaugh This is pretty interesting. Our last project really showed how difficult it is estimating good parameters for oscillating data. Also, I should really start looking into python more...it can do a lot of cool things.

[Carlos Valdez @c_eleazarV]

@austinharper59 Cool. This will probably be a good reference for later.

[Carlos Valdez @c_eleazarV]

Since we didn't get to talk a lot about Singular Value Decomposition, here's an example of how to do it
https://t.co/50FrenQXvD
#math3185

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB It makes a good point about how a high coefficient might not mean the model is actually good or vise versa. I've seen in our last project how one model has less residual error than another but in the long run will be a worse model.

[Carlos Valdez @c_eleazarV]

@austinharper59 Interesting! I remember writing a forum post over simulated annealing which is a type of metaheuristic optimization. It finds approximate global optimums rather than precise local optimums.

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB @holston_sebaugh What makes real world problems so good is that it helps understand concepts such as constraints. By just thinking about numbers you may be like, ok x,y>0 whatever. It's not until you see it's application and then you're like, oh yeah you can't have a negative amount of milk

[Carlos Valdez @c_eleazarV]

This article discusses the R squared coefficient and how to interpret it.
https://t.co/t3K5CZWF01
#math3185

[Carlos Valdez @c_eleazarV]

@CamPerkinsQB Great find Cam! This is a good example of problem solving in the real world...being able to take a problem and expressing it in mathematical terms. I'll keep this example for future reference as we study linear programming.

[Carlos Valdez @c_eleazarV]

@holston_sebaugh This gives a nice explanation. I like being able to see python code because me helps me understand the coding concepts since I'm not as familiar with python

[Carlos Valdez @c_eleazarV]

@austinharper59 I like the fact that the correlation coefficient is the slope. Spreading out the x-values causes the slope to be more horizontal and spreading out the y-values causes the slope to be more vertical. It makes sense!