Ever wondered how I consume research so fast? I'm going to describe the process i use to read lots of machine learning research papers fast and efficiently. It's basically a 3-pass approach, i'll go over the details and show you the extra resources I use to learn these advanced topics. You don't have to be a PhD, anyone can read research papers. It just takes practice and patience.
Please Subscribe! And like. And comment. That's what keeps me going.
Want more education? Connect with me here:
More learning resources:
Join us in the Wizards Slack channel:
And please support me on Patreon:
Signup for my newsletter for exciting updates in the field of AI:
I’m a masters student in molecular bio and I have to say that they don’t really teach us hoe to read/navigate research articles (at least mine doesn’t) it’s a learn as you go .. but your video was very helpful thank you.
I have just started my PhD journey, and I have a question in regards to reading an article for the first pass to find out the aim of the paper, then you said you will look through the internet why is this paper is popular or what other people are saying about it, so how do you find out what other people are saying about a certain published article?
This are very useful tips. Will help me go through all of them long university research papers.
Also, I did a video about how to write a research paper... feel free to check it out in my channel... could be helpful to you when you will start writing your own research papers :))
This was useful but too high level. You ought to have walked us through an actual example of understanding parts of a papers, leading to understanding the whole. You could of has picked a simple 2-page paper as a sample.
I am failing my Statistics and I need to do a research paper to get at least 3points to get a C. Can u help me to do my paper? I am computer illiterate so I do not know how to do graphs and pie charts and all....even U have a concept for my research paper...please help. TQ.
you mentioned Arxiv Sanity Preserver ... great concept. Here is one more I use that I find helpful if I want to dig deeper into the paper. Bring it up in Google Scholar and look at who cited it. There is a direct correlation between influential/importance/seminal and number of folks who cited it. Keep on truckin'
thanks for the insights ... I use a similar method, except I work in two modes : goal directed and canvass mode (quasi-random). Which mode I use depends on what part of my brain has decided to work well. What the latter mode allows me to hopefully do is "run" into that strange, quirky paper that generates "out-of-the-box" ideas. I usually will find those papers in the application of AI, Machine Learning or Computer Vision. E.g. Biomedical fields, or Plant Science. Currently I am working on a concept : images as a language of textures (great approach for dealing with occlusion). Which led me to Oxford Robotics (textures) and CMU Graph Learning. Best in everything.
Hey siraj.. thanks for your continuous effort for letting us know about different techniques via available technologies. I did not really get the concept of getting the relevant paper code, how to get, could you help?
Have a goal (something to build). This will guide your reading towards what is actually practical and relevant.
1. Skim. Title, Abstract, (Why should I care? What problems does this technique solve? Is it worth my time to keep analyzing this paper?)
2. Take notes on all the English, plots and graphs, look for stuff he already understands so he can put it in context. Read code and understand it through comments. Teaching it helps me (Siraj) understand it.
a. Explain it on YouTube to truly understand all the details.
3. Math: Write out derivations if you do this! Don't just read it. He'll recode it if it's really important to do so; this gives him a thorough understanding of what the tech does. Then he'll discuss with others to make sure he actually understands what it does
If necessary, study the math on Khan Academy
Ask for help! (stackoverflow.com, machinelearningmastery.com, etc.)
In case you have your dataset and you wants to import it via pandas or numpy using import function, do you need to save your data into numpy library or? If yes how do you do this?
I have a dataset from internet Im trying to practice to manipulate the data using numpy import function, but it tells me file error. Im suggesting that maybe i need to save the file in the desktop and if i go to python and use import function it doesn't work.. Please i need explanation how to solve this...
You may reach me via email [email protected]
Your aid is highly needed thank you.
Im thinking how do i need to save the file for it to be recognise as by numpy import function.
This approach works quite well for many ML papers, especially since there is a lot of "empirical validation" papers coming out right now; I also include figures on the first pass.
However, for more theoretical papers I find it better to read the equations with the English text, as they usually refer to one another and you can't make sense of anything otherwise.
That said, I do skip the Methods and implementation details entirely IF the paper passed peer-review (I assume that the peer was good enough to spot flaws) and only really bother with that if I actually need it for my work (e.g. I want to cite it or base my research off it).
Hey Siraj! I can't thank you enough brother for those resources links to the awesome ML/Data papers. I wish if the majority of Professors in the Universities were at least half as interesting/motivating as you. Keep up the great work!
Tax Deposit Information For the date range, provides taxable earnings totals, Federal 941, 940 and State tax deposit information.
This item has been saved to your reading list.
The level of manual tasks, duplication and non-value added activities in the payroll process often prevent staff from providing the reports that management require on a timely basis.
The best practice for providing management information that is timely, useful and relevant is automating, simplifying and standardising as many processes and tasks as possible, to eliminate unnecessary complexity and inconsistencies. This will reduce opportunities for error or fraud and make it easier to track the quantity and quality of information flowing to payroll.
A holistic approach that considers factors across all business dimensions, including strategy, structure, people, process and technology, creates reporting that is highly sustainable and scalable.
If you do not feel confident in your reporting capabilities or would like to improve your reports, technology could be a valuable ally providing master data that is clean, up to date and which can be relied upon.
Employee Advance Summary For the date range lists Advances paid and repaid with beginning and ending balances, one line per employee.
Employee Payroll Summary Lists for the date range Gross Pay, Total Withholdings, Credits, Reimbursements, Advances and Checks Total, one employee per line.
Payroll Journal Summary.
Payroll Journal Summary For the date range selected provides posting debit and credit totals for the Accounting sections Assets, Liabilities, Revenue and Expenses, with totals by General Ledger account number within each section.
Tax Liability Summary.
Form 941 Information For the date range selected, provides total Federal and State taxable wages, EIC credits, and other tax deposit information relevant to the 941 report.