Difficult:
I would like to understand why the unbolded w to the power of (k-m)*l is 0 whenever k != m. The calculation is just a summation which I would expect to have a value not equal to 0.
Interesting:
I just computed the matrix W that is shown in algorithm 8.4.1 for the lab. This way of computing the discrete Fourier transforms is a lot quicker than doing a loop.
Filtering noise is a very cool and useful application of Fourier Tranforms. It is interesting that the filtering method that does not focus on the desired frequency range is the one that matches the data the least. I would have expected it to be the other way around.
Thursday, 29 November 2018
Tuesday, 27 November 2018
Section 8.3: Convergence of Fourier Series
Difficult:
I followed most lemmata and saw how the logic behind the steps. I wish the chapter would have had more application/motivation because I don't really know why we are proving all these things. For example, the last proof that was about a theorem in the first subsection of chapter 8, I don't even know the purpose of that theorem.
Interesting:
i) The graph is cool. It was the highlight, which says more about the this chapter than I have to say with words
ii) the plural of lemma is lemmata
I followed most lemmata and saw how the logic behind the steps. I wish the chapter would have had more application/motivation because I don't really know why we are proving all these things. For example, the last proof that was about a theorem in the first subsection of chapter 8, I don't even know the purpose of that theorem.
Interesting:
i) The graph is cool. It was the highlight, which says more about the this chapter than I have to say with words
ii) the plural of lemma is lemmata
Sunday, 25 November 2018
Section 8.1: Fourier Series. & Appendix B
Difficult:
There was a lot of new information in this reading assignment. It is just hard to know if I will need to solve the problems out of what I read. I think if there was more visual representation of the projections that are happening in the equation it would become more intuitive to use them.
Insightful:
I am looking forward to learning how to use Fourier Series since I've heard it is used extensively in engineering fields. The part on number music theory was also interesting but I am looking to use it on what interests me.
There was a lot of new information in this reading assignment. It is just hard to know if I will need to solve the problems out of what I read. I think if there was more visual representation of the projections that are happening in the equation it would become more intuitive to use them.
Insightful:
I am looking forward to learning how to use Fourier Series since I've heard it is used extensively in engineering fields. The part on number music theory was also interesting but I am looking to use it on what interests me.
Tuesday, 20 November 2018
Section 7.3: Hashing
Difficult:
Conceptually, everything made sense in this section. The only difficult thing I see is deciding which method to use for handling collisions. They all have a potential con but it really does depend on the type of data you are storing.
Insightful:
I like the fact that the probability for having the same birthday came back for something more practical. I also like what we are learning because when I learned that a set had a hash function in it made me want to use sets more in my code. Finding out the usefulness of something can help us get away from comfortable habits.
Conceptually, everything made sense in this section. The only difficult thing I see is deciding which method to use for handling collisions. They all have a potential con but it really does depend on the type of data you are storing.
Insightful:
I like the fact that the probability for having the same birthday came back for something more practical. I also like what we are learning because when I learned that a set had a hash function in it made me want to use sets more in my code. Finding out the usefulness of something can help us get away from comfortable habits.
Sunday, 18 November 2018
Section 7.2: Importance, Inversion, and Rejection Sampling
Difficult:
In this section, we are learning about using different distributions to use the Monte Carlo Methods. We are trying to match it as much as possible to a function that we are using Monte Carlo on. I want to understand the importance of the inverse sampling better because I feel like I did not understand the examples at the end of the subsection.
Interesting:
Rejection Sampling uses two methods. One that can be more accurate but can also diverge. If the first diverges the second, the more reliable method is used. Pretty much an if statement that can help with the approximation of the model.
In this section, we are learning about using different distributions to use the Monte Carlo Methods. We are trying to match it as much as possible to a function that we are using Monte Carlo on. I want to understand the importance of the inverse sampling better because I feel like I did not understand the examples at the end of the subsection.
Interesting:
Rejection Sampling uses two methods. One that can be more accurate but can also diverge. If the first diverges the second, the more reliable method is used. Pretty much an if statement that can help with the approximation of the model.
Thursday, 15 November 2018
Section 7.1: Monte Carlo Methods
Difficult:
I don't think there was anything too difficult about this section but that does not mean I would not have a hard time explaining the concept. The main idea I am getting out of it is that for the outcomes of random variables in a distribution we can compute the expected value of integrals and other models. Like the Riemann sum example where the same amount of points were taken and because of the distribution, we get a pretty good estimation of the integral itself. I am excited to learn about these methods because I just went to a meeting today where this idea came up.
Insightful:
The approximation of pi is remarkable. It is mind boggling how random selected values can give such an accurate response.
I don't think there was anything too difficult about this section but that does not mean I would not have a hard time explaining the concept. The main idea I am getting out of it is that for the outcomes of random variables in a distribution we can compute the expected value of integrals and other models. Like the Riemann sum example where the same amount of points were taken and because of the distribution, we get a pretty good estimation of the integral itself. I am excited to learn about these methods because I just went to a meeting today where this idea came up.
Insightful:
The approximation of pi is remarkable. It is mind boggling how random selected values can give such an accurate response.
Tuesday, 13 November 2018
Test Prep
- Which topics and ideas do you think are the most important out of those we have studied?
- What kinds of questions do you expect to see on the exam?
Calculating probabilities with Bayes' rule
Doing lots of algebra
- What do you need to work on understanding better before the exam?
Saturday, 10 November 2018
Section 6.5: Bayesian Statistics
Difficult:
The unknown paramenters are still a concept I am getting used to. Condition probability with many paramenters and changing outcomes is the idea I am trying to rap my mind around. Fortunately, this section has examples to help with that.
Insightful:
I find it interesting that you can mix two different types of distributions and get a similar or different distribution out. The idea of which one to use is still a little hazy but nothing practice can't teach.
The unknown paramenters are still a concept I am getting used to. Condition probability with many paramenters and changing outcomes is the idea I am trying to rap my mind around. Fortunately, this section has examples to help with that.
Insightful:
I find it interesting that you can mix two different types of distributions and get a similar or different distribution out. The idea of which one to use is still a little hazy but nothing practice can't teach.
Thursday, 8 November 2018
Section 6.3: The Central Limit Theorem
Difficult:
How come in the beta distribution graphs the more likely outcome to the left did not happen more often? I am now assuming the scripted N is the normal distribution simulated by the Central Limit Theorem (since it was not defined).
Insightful:
I don't understand notation but I understand the idea. If we have events happening a lot, they start to simulate a normal distribution. Even if they were drawn from a different distribution.
How come in the beta distribution graphs the more likely outcome to the left did not happen more often? I am now assuming the scripted N is the normal distribution simulated by the Central Limit Theorem (since it was not defined).
Insightful:
I don't understand notation but I understand the idea. If we have events happening a lot, they start to simulate a normal distribution. Even if they were drawn from a different distribution.
Tuesday, 6 November 2018
Section 6:2 The Law of Large Numbers
Difficult:
The theorems in this chapter are fairly straightforward but I am afraid that I have forgotten parts of the last chapter that are used as facts in this chapter. It is great to review but at the same time I feel like there are concepts going over my head even though I am supposed to be understanding them from previous material (for example the variable 1E) . From now on, I will review the previous chapters as I do the homework.
Insightful:
I am starting to see from the material in this call and the volume 1 class that inequalities will be an important part of problem-solving. I can see the bounds on how probable an event is as useful information when making a decision (whether it be business or engineering).
The theorems in this chapter are fairly straightforward but I am afraid that I have forgotten parts of the last chapter that are used as facts in this chapter. It is great to review but at the same time I feel like there are concepts going over my head even though I am supposed to be understanding them from previous material (for example the variable 1E) . From now on, I will review the previous chapters as I do the homework.
Insightful:
I am starting to see from the material in this call and the volume 1 class that inequalities will be an important part of problem-solving. I can see the bounds on how probable an event is as useful information when making a decision (whether it be business or engineering).
Sunday, 4 November 2018
Section 6.1: Estimation
Difficult:
The book could have done a better job at defining certain terms like parameters, or likeliness. The examples seemed more like proofs than real-life examples. I was trying to make my own examples to try to understand the material better but I don't know if my assumptions are correct.
Insightful:
I like the fact that chapter 6 is a continuation of chapter 5 since I feel like more exposure to these terms will make them stick better. Hopefully, everything will make more sense as we get closer to the test.
The book could have done a better job at defining certain terms like parameters, or likeliness. The examples seemed more like proofs than real-life examples. I was trying to make my own examples to try to understand the material better but I don't know if my assumptions are correct.
Insightful:
I like the fact that chapter 6 is a continuation of chapter 5 since I feel like more exposure to these terms will make them stick better. Hopefully, everything will make more sense as we get closer to the test.
Thursday, 1 November 2018
Section 5.7: Multivariate Radom Variables
Difficult:
Everything. This chapter was extremely dense. There is multiply options in a probability space. But the logic for the proves was not easy to follow
Questionable:
Why is the expected value greater than 1 for the different options of the balls example?
Everything. This chapter was extremely dense. There is multiply options in a probability space. But the logic for the proves was not easy to follow
Questionable:
Why is the expected value greater than 1 for the different options of the balls example?
Subscribe to:
Posts (Atom)