MATH 151A
Applied Numerical Methods
Description: Lecture, three hours; discussion, one hour. Requisites: courses 32B, 33B, 115A, Program in Computing 10A or Computer Science 31. Introduction to numerical methods with emphasis on algorithms, analysis of algorithms, and computer implementation issues. Solution of nonlinear equations. Numerical differentiation, integration, and interpolation. Direct methods for solving linear systems. Letter grading.
Units: 4.0
Units: 4.0
2 of 5
Most Helpful Review
Spring 2018 - Okay. I'll speak from personal experience here. This is for 151B by the way. In a nutshell, Deutsch's lectures aren't the best, but the graded classwork and exams are very manageable. Lecture-wise, his handwriting was hard to read, and he would repeat what he writes without giving much context. Thus, it's super hard to digest the information, especially with the very algorithmic and computational nature expected of the material in 151B. However, if you need to take the class, what I advise is to truly take some time to sit through and understand the corresponding sections in the book deeply. His tests do not have unreasonably hard curveball questions, so you should end up thoroughly knowing how to do the easier derivations and how each method works, as well as advantages/disadvantages. I mean, I could tell that he's a nice professor too, as he offered to drop one of our midterms later in the quarter and curved pretty generously at >=85% being at least an A- (probably because the final had a median of 70, but in my opinion, it was pretty fair). All in all, unclear lectures, but manageable work and tests.
Spring 2018 - Okay. I'll speak from personal experience here. This is for 151B by the way. In a nutshell, Deutsch's lectures aren't the best, but the graded classwork and exams are very manageable. Lecture-wise, his handwriting was hard to read, and he would repeat what he writes without giving much context. Thus, it's super hard to digest the information, especially with the very algorithmic and computational nature expected of the material in 151B. However, if you need to take the class, what I advise is to truly take some time to sit through and understand the corresponding sections in the book deeply. His tests do not have unreasonably hard curveball questions, so you should end up thoroughly knowing how to do the easier derivations and how each method works, as well as advantages/disadvantages. I mean, I could tell that he's a nice professor too, as he offered to drop one of our midterms later in the quarter and curved pretty generously at >=85% being at least an A- (probably because the final had a median of 70, but in my opinion, it was pretty fair). All in all, unclear lectures, but manageable work and tests.
Most Helpful Review
I highly recommend Professor DeVita! He is a pretty young guy, especially compared to others in the math dept, so he can relate to undergrads. Upper-divs can be a great or horrible experience, depending on the professor. With DeVita you will get effective lectures that go by the book, concise homework, and very fair exams (open book/notes!!). I was in 151A the term before with Fattorhini, so I can directly compare their teaching styles. Even though Fattorhini was very friendly and wanted us to learn, it was clear that he cared more about theory than the application or large-scale computing. Even though it was supposed to be numerical ANALYSIS, that professor focused on proofs and barely touched coding. The difference with DeVita was night-and-day; he approached the same topics with a focus on application and encouraged use of programs like Matlab. The pace was great for an intro class.. he started with the basics of using a programming language and familiar processes like bisection method. Lectures were very clear and involved both derivations and examples.. no sweat if you missed lecture, because it was clear which section in the book to read. My favorite thing about Professor DeVita was his awareness of industry application. A lot of math professors forget that most undergrads are more interested in industry than academia. Professor DeVita spent just as much time discussing effective coding (minimizing computational error, using binary, etc) as he did on theorem proofs. I look forward to taking 151B with him next quarter :)
I highly recommend Professor DeVita! He is a pretty young guy, especially compared to others in the math dept, so he can relate to undergrads. Upper-divs can be a great or horrible experience, depending on the professor. With DeVita you will get effective lectures that go by the book, concise homework, and very fair exams (open book/notes!!). I was in 151A the term before with Fattorhini, so I can directly compare their teaching styles. Even though Fattorhini was very friendly and wanted us to learn, it was clear that he cared more about theory than the application or large-scale computing. Even though it was supposed to be numerical ANALYSIS, that professor focused on proofs and barely touched coding. The difference with DeVita was night-and-day; he approached the same topics with a focus on application and encouraged use of programs like Matlab. The pace was great for an intro class.. he started with the basics of using a programming language and familiar processes like bisection method. Lectures were very clear and involved both derivations and examples.. no sweat if you missed lecture, because it was clear which section in the book to read. My favorite thing about Professor DeVita was his awareness of industry application. A lot of math professors forget that most undergrads are more interested in industry than academia. Professor DeVita spent just as much time discussing effective coding (minimizing computational error, using binary, etc) as he did on theorem proofs. I look forward to taking 151B with him next quarter :)
AD
Most Helpful Review
Winter 2021 - NOT recommended. The grading of the class is based on homework (from textbook), projects (coding with Matlab and report in Latex), two midterms, and a final project (code + report + slides + live presentation and Q&A with professor). Overall: I personally learn little from this class. The professor is a nice guy but his handwriting is hard to read even in remote instruction, and I feel like he's not very good at explaining things. He really wanted interactions but sadly people don't always respond and it's kinda awkward. Homework assignments: The homework is unbelievably long.. It's not that there are a large number of questions, it's just the calculation sucks. While I understand this is a numerical method class and some amount of calculation is necessary, this is definitely too much- I'm not a computer and it sucks to calculate a system of 20+ equations, etc. Huge amount of repeated exercises with LOTS OF calculations makes the HW's such a pain. And it sucks oven more when the last homework is released on Mon of finals week and due on Friday.. Just imagine you climb out of your sofa after finally finished all your finals to do the homework with ton of calculations.. PAIN Projects: Honestly I'd say projects are ok mainly because they are graded kindly, despite the spec is unclear. You might get confused on what the professor really wants, but it's usually OK to just make an assumption and go on. So doable projects even with no previous experience in Matlab! Exams: Doable. Mainly about concepts- it makes me feel better since at least the exams are not THAT calculation intensive. But something to watch out: the schedule of midterms are pretty randomly decided and you'll not get an official notification until the night before (though you might know the day of midterm via discussion forums).
Winter 2021 - NOT recommended. The grading of the class is based on homework (from textbook), projects (coding with Matlab and report in Latex), two midterms, and a final project (code + report + slides + live presentation and Q&A with professor). Overall: I personally learn little from this class. The professor is a nice guy but his handwriting is hard to read even in remote instruction, and I feel like he's not very good at explaining things. He really wanted interactions but sadly people don't always respond and it's kinda awkward. Homework assignments: The homework is unbelievably long.. It's not that there are a large number of questions, it's just the calculation sucks. While I understand this is a numerical method class and some amount of calculation is necessary, this is definitely too much- I'm not a computer and it sucks to calculate a system of 20+ equations, etc. Huge amount of repeated exercises with LOTS OF calculations makes the HW's such a pain. And it sucks oven more when the last homework is released on Mon of finals week and due on Friday.. Just imagine you climb out of your sofa after finally finished all your finals to do the homework with ton of calculations.. PAIN Projects: Honestly I'd say projects are ok mainly because they are graded kindly, despite the spec is unclear. You might get confused on what the professor really wants, but it's usually OK to just make an assumption and go on. So doable projects even with no previous experience in Matlab! Exams: Doable. Mainly about concepts- it makes me feel better since at least the exams are not THAT calculation intensive. But something to watch out: the schedule of midterms are pretty randomly decided and you'll not get an official notification until the night before (though you might know the day of midterm via discussion forums).
AD
Most Helpful Review
Winter 2024 - The professor is extremely nice and helpful. His lecture notes are very clear, so I was always able to follow along. The pace of the course is reasonable, and I am able to strengthen my understanding of the concepts through working on HW problems, which prompt you to think and practice but are not too challenging to the point that you just have no idea how to solve them. The midterm and final tests are also very accommodating, where we had open book and notes and questions that focus more on implementing the numerical methods rather than on the very theoretical stuff (at least for the midterm). One thing I'm not extremely comfortable with is using Matlab because I've never really used it before, but it should still be ok to handle if you have some programming experience before. Overall, I would really recommend taking the course with Professor Chenfanfu Jiang!
Winter 2024 - The professor is extremely nice and helpful. His lecture notes are very clear, so I was always able to follow along. The pace of the course is reasonable, and I am able to strengthen my understanding of the concepts through working on HW problems, which prompt you to think and practice but are not too challenging to the point that you just have no idea how to solve them. The midterm and final tests are also very accommodating, where we had open book and notes and questions that focus more on implementing the numerical methods rather than on the very theoretical stuff (at least for the midterm). One thing I'm not extremely comfortable with is using Matlab because I've never really used it before, but it should still be ok to handle if you have some programming experience before. Overall, I would really recommend taking the course with Professor Chenfanfu Jiang!
2 of 5
