|
|
| Line 1: |
Line 1: |
| − | This is a preparatory course for the main part of My first scientific paper. Its goal is to distribute works of scientific research evenly over a year.
| + | == The seminar == |
| − | | + | # [https://m1p.org/index.php?title=Step_1 The warm-up 3-minute test] |
| − | Scientific research is a collective activity, and your main goal during this course is to find a scientific advisor who devotes their time to you in exchange for ''academic'' results. So, in the end, you have some skills in how to
| + | # Model, Algorithm, Method: Machine learning in a nut-shell |
| − | # select a research topic, | + | <!-- #* more terms: statistical hypothesis, algebraic structure, model selection, bayesian inference --> |
| − | # critically analyze the literature, | + | # Step 0 homework results discussion |
| − | # state the problem, | + | # Step 1 homework, how to read (how to search is a separate topic) |
| − | # and convey your vell-reasoned message to the reader. | + | # Structure of the main message |
| − | | + | # Structure of the abstract |
| − | But again, '''your main goal''' is to find a highly qualified scientific adviser (with their team) who gifts you valuable time.
| + | # Extracting keywords |
| − | | + | # Highlights: compressing the paper |
| − | To accomplish the homework of this course, you may select a topic in applied mathematics or theoretical machine learning. Or better and we recommend it, you can change your topic after each step so you can feel various fields of your future expertise.
| + | # Instastructure for your homework |
| | + | #* GitHub: organize the repository |
| | + | #* LaTeX: compile your file and commit without temporary files |
| | + | # The papers to select from |
| | + | # Optional GPT-role discussion |
| | | | |
| | ==Resources== | | ==Resources== |
| Line 20: |
Line 24: |
| | | | |
| | ==Homework== | | ==Homework== |
| − | # [https://forms.gle/y6kbSz7wLX91stZ19 Fill out the form to attend the course]. Keep your email to log in. | + | # Set up your GitHub repository using the template |
| − | # Set up the tools | + | # Select a paper to read from the list below |
| − | ## Online LaTex is [https://www.overleaf.com/ Overleaf] | + | # Write your own |
| − | ## Offline LaTex is [https://miktex.org/ MikTex] with [https://www.winedt.com/ WinEdt] or [https://pages.uoregon.edu/koch/texshop TexShop] with [https://www.xm1math.net/texmaker/ TeXMaker] | + | ## Abstract |
| − | ## Offline BibTex is [https://www.jabref.org JabRef] | + | ## Keywords |
| − | ## Sign up for [https://github.com GitHub] to keep your progress | + | ## Highlights |
| − | # If you are not familiar with the [https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf C.P. Bishop's book, 2006] or 2024, start reading | + | ## Short motivation for why you selected this paper (no templates here it is an extra topic to discuss) |
| − | # Watch a useful lecture [https://www.youtube.com/watch?v=Unzc731iCUY "How to speak"] by Patrick Winston, 2018 | + | # Compile and upload TEX and PDF to GitHub (no temporary files, please) |
| | + | # Linear models. Read pages XX from the book XX, for the next warm-up test |
| | | | |
| | + | ==Papers to read== |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | # |
| | + | |
| | + | Can I select any paper from the internet by my own choice? – Yes. Here are the formal requirements. |
| | + | # A clear message in the area of Theoretical Foundations of Machine Learning. |
| | + | # Top peer review journals, no ArXiv, better avoid conferences. |
| | + | ## |
| | + | ## |
| | + | ## |
| | + | ## |
| | + | ## |
| | + | ## |
| | + | ## |
| | + | # No overviews, it is another genre. |
| | + | # No Kaggle-style papers with messages like "It works, but nobody knows how". |
| | + | # No papers from another field: Linguistics, Medicine, Finance, Physics, etc. There must be only one primary subject: Machine Learning. |
| | + | |
| | + | |
| | + | <!-- |
| | ==Fun== | | ==Fun== |
| − | Why do we need weekly homework? Look [https://www.youtube.com/shorts/Rvmvt7gscIM how two neurons connect one another]. So we learn as we train. <!-- [https://www.youtube.com/watch?v=ehbFoALnV4o how a chick's neurons develop] connecting the opposite side of the central nervous system. -->
| + | To do)) |
| | + | --> |
| | | | |
| | ==Transcript of the video== | | ==Transcript of the video== |
| − | Hello dear colleagues!
| + | Appears after the seminar. |
| − | | |
| − | This is an orientation hour for The Art of Scientific Research.
| |
| − | It is a preparatory course for your thesis in Applied Mathematics and Machine Learning.
| |
| − | | |
| − | You want to
| |
| − | 1) fit research activity,
| |
| − | 2) be part of the scientific community,
| |
| − | 3) define your field of expertise.
| |
| − | You aim to find your scientific advisor and carefully select your research project.
| |
| − | | |
| − | The task of project selection could be more difficult than the task of project accomplishment since it requires higher qualifications from the researcher. So, we split the whole thing into two parts. The first one is The Art. The second one is called My First Scientific Paper. The second one is already more than 10 years old. It has a separate schedule, which starts the next spring. It is already well established, unlike this new course so it is an experimental one. The organizers' goal is to boost the quality of student thesis works.
| |
| − | | |
| − | This student's works highly depend on how the student fits the scientific community and this kind of community must be organized in the following way. First, a student is a project driver, highly committed to their activity. Second, a consultant (usually it is a graduated student or um PhD student) who (an hour a week) helps the younger student. The third one is a professor an expert in the field who is responsible for the end result of the project. In the beginning, this professor states the problem and in the end, harvests the results.
| |
| − | | |
| − | The production of scientific results and reporting it is our ultimate goal. To do it we have to learn
| |
| − | how to state the problem
| |
| − | how to recognize if the project is feasible and
| |
| − | how to present our results
| |
| − | We organize each seminar in the following way.
| |
| − | First, there will be a minute test and a brief analysis.
| |
| − | Second, the theoretical part of the theory will be about the style of scientific research and about some aspects of machine learning.
| |
| − | After we talk about the homework and then discuss the homework we have done.
| |
| − | Each semester comprises two modules. The first will be about how to deliver your message how to pitch your project, and the second one is about how to reason your project, and how to establish the theoretical part of it.
| |
| − | | |
| − | You can fix your favorite project in the beginning but I recommend you change your subject of research each week to feel different subjects.
| |
| − | This plan of classes is provisory but anyhow it coincides with the road map of a scientific paper preparation we select a paper to start from, discuss its principles, formulate results, collect a review, and prepare a small talk about it.
| |
| − | | |
| − | Here we walked through some themes to discuss. I hope it shrinks. Each our seminar will have a short theoretical part. But it will not introduce the methods themselves. It will introduce how to deliver your message to your reader and your audience in an easy and fast way.
| |
| − | | |
| − | The scoring. We would like to avoid the written exam, so there will be weekly scoring with tests at the beginning of a seminar, your talks at the end of a seminar, and your written homework.
| |
| − | Of course, each module ends with a small course work and the first module is about ones-slide talk. Plus about two pages of project description. So each week you will get several points these points add up and are scaled. The deadlines are strict.
| |
| − | | |
| − | Since this course is a new one, we don't know how it ends. I hope everything will be fine in the case of a large group of students there will be no possibility of reviewing all your texts and listening to all your talks. So the feedback will be limited.
| |
| − | | |
| − | It is expected that you have a bachelor's degree. But these first six items are about just the first two years of bachelor's study. And the last seventh item is about your basic knowledge of machine learning. We recommend the book of Christopher Bishop as a standard of machine learning knowledge.
| |
| − | Here are the main references but in fact, these references will be spread over the homework and some of these books are quite large to read. So we will just point to their parts.
| |
| − | Is expected to be Saturday 2:40 p.m.
| |
| − | | |
| − | We go to the Step Zero.
| |
| − | Please read this motivational text, read about the syllabus of this part, and the part of my first scientific paper.
| |
| − | | |
| − | The will be a link to this video. I put a slow-speed version of this talk here.
| |
| − | Despite the fact I mentioned this is the orientation hour there will be homework.
| |
| − | First of all please subscribe to attend to the course. Fill out the form and the form will collect your email this email will be your ID. Please do not change it unless you want to lose your course scores. The deadline is late it's September 20. You will have some time to think about your commitment.
| |
| − | Your name how to address to you, University student group.
| |
| − | | |
| − | To engage yourself please think about your professional goal in connection with your future thesis.
| |
| − | The required question why would you want to take this course?
| |
| − | What is the profit in it for you?
| |
| − | | |
| − | You answer this and this will be the first homework.
| |
| − | Don't forget to press the submit button.
| |
| − | | |
| − | Also to keep your time in the future homework please set up some tools.
| |
| − | If you are not familiar with LaTeX, please find a crash course on this and start reading it.
| |
| − | We will keep our progress in two archives. The first one is the Google forms and the second one is your GitHub repository.
| |
| − | | |
| − | Also if you're not familiar with the Bishop's book just start reading it and
| |
| − | For fun and for homework watch a lecture by Patrick Winston from MIT on how to speak. We will talk about it in the next class.
| |
| − | | |
| − | So fill out your questionnaire and see you next week.
| |
| − | | |
| − | | |
| − | | |
| − | Hey, the test will be about Patrick Winston's lecture.
| |