Hi Frank! Thanks for the comment. I like this idea of showing how to make nice figures for papers. I usually export my figures in SVG format and then edit them using software like CorelDRAW. Will try to make a video about this!
I feel that most of things stated in this video can be summed up to couple general points: 1. Increase ease of information consumption. 2. Increase density of information. Pretty much all topics you cover do fit both of these statements, but use obscure terms like "people don't like reading", "useless diagrams", etc. People who read papers do not mind reading, but don't like reading text that is off-topic. People do like pictures that simplify material understanding, not ones that take space. This is the same issue one can encounter not only when writing papers, but also do most of paper work at any (in my case tech) organization. Writing design docs, instructions, user documentation, etc. One of the best ways for learning to write better documentation I found is: read documentation you wrote couple months ago and forgot everything about. If you can write it better now, you learned something. Unfortunately, it is really hard to estimate both of given metrics while working on paper due to difference of perception. People that will read your paper most likely know nothing about background and details of the topic. While you, as the author, can use multiple assumptions that seem obvious at the moment or add details that seem close to your heart, but not important for the reader.
Andre, you are doing great. keep up the good work.. also make a video for fresh PhD jobs options.. and also a specific video on life in your university, also a video on your daily routine in university - Very Good Luck.
A very nice video. I am currently a graduate student and would like to know, just in your opinion which softwares are the best for making different figures. Also maybe you can explain what kind of style should be followed in particular for each figure. Perhaps you could put some information about the these aspects in your next video. Thanks
Hi! Thanks for the comment. Yes, I plan to make a video about figures for papers. In short, you need to produce vector-based figures (say, in SVG or PDF) from the data you have. You can do this in any software or programming language, for example, plotting packages in Python, Julia, Matlab. Then, you can further edit and improve these figures using vector graphics software (Adobe Illustrator, CorelDRAW, etc.). This is a long creative process where you check if you like the colours, font size and style, alignment of different parts of the figures, and so on. Finally, import the vector-base images into your paper, that is, try to avoid any pixelated figures.
Is the literature "radar" you showed at 15:54 original or it's common in your field because this is the first time i see something like it and I'd love to use it in my work, it's way better than a complex table.Also, how did you make it ?
Hi! No, I don't think literature radars are common in my field. It was my initiative to present the literature review in an unusual way. I drew that radar myself in CorelDRAW. You can use other similar software like Inkscape or Adobe Illustrator. Basically, you create a circle, split it into parts (research directions) and manually place papers that you want to mention.
Will be great if u do new video about the source code . For my failed im doing my PhD with communication and information engineering and my research about Mobile edge computing enabled UAV . And how can i get the source code and how basically can code the paper and if i planned to code the paper by myself will led to the results the author shows! I mean optimization problem and mostly using convex optimization . I’m struggling with this part
Hi Hussein! Thanks for the comments. Do you mean developing a code to get results for a paper? I am not sure what coding you are dealing with. If speaking about optimization (say, coding mathematical optimization models in Python or Julia), there are many videos/tutorials on RUclips. Did you see optimization webinars by Gurobi? They have some very basic and clear explanations. Also, this article is useful: www.gurobi.com/resources/mixed-integer-programming-mip-a-primer-on-the-basics/
Minkowski addition to studying the capabilities of Virtual Power Plants is business as usual today, by the way, people who are not in the field still have problems catching this concept.
Hi! Yes, conference papers tend to be more lenient. This is especially true for student conferences (where the majority of submissions come from MSc and PhD students). At such conferences, they decline papers only if they are complete nonsense. However, there are some tough conferences where they expect each paper to be novel and professional. I know a few in my field (e.g., Power Systems Computation Conference) and have heard that the ML community has conferences with rigorous requirements. It is better to check the requirements and acceptance statistics in advance. Good luck!
Hi! The short answer is: I am using Plots.jl package for Julia programming language; I then save figures in SVG or PDF formats and fine-tune them using CorelDRAW software. I plan to publish a video tutorial about that (how I create figures) in a month or two. So stay tuned! Andrey
I just love the metaphor at the end between videos on RUclips and papers. They have the same goal to be watched and shared !
This video was extremely helpful for me. Could you make a video about the process of making your scientific graphics?
Hi Frank! Thanks for the comment. I like this idea of showing how to make nice figures for papers. I usually export my figures in SVG format and then edit them using software like CorelDRAW. Will try to make a video about this!
Thanks for your advice. It motivates me to continue working on my script!
I feel that most of things stated in this video can be summed up to couple general points:
1. Increase ease of information consumption.
2. Increase density of information.
Pretty much all topics you cover do fit both of these statements, but use obscure terms like "people don't like reading", "useless diagrams", etc.
People who read papers do not mind reading, but don't like reading text that is off-topic. People do like pictures that simplify material understanding, not ones that take space.
This is the same issue one can encounter not only when writing papers, but also do most of paper work at any (in my case tech) organization. Writing design docs, instructions, user documentation, etc.
One of the best ways for learning to write better documentation I found is: read documentation you wrote couple months ago and forgot everything about. If you can write it better now, you learned something.
Unfortunately, it is really hard to estimate both of given metrics while working on paper due to difference of perception. People that will read your paper most likely know nothing about background and details of the topic. While you, as the author, can use multiple assumptions that seem obvious at the moment or add details that seem close to your heart, but not important for the reader.
And thank u so much for this video . Great advices and so informative
This is very informative. More videos, please.
Andre, you are doing great. keep up the good work.. also make a video for fresh PhD jobs options.. and also a specific video on life in your university, also a video on your daily routine in university - Very Good Luck.
Nice work; best wishes to you.
Great explanations Andrey 👍🏾
Thank you for the High-quality video.Keep going
Thank you, Andrey!
A very nice video. I am currently a graduate student and would like to know, just in your opinion which softwares are the best for making different figures. Also maybe you can explain what kind of style should be followed in particular for each figure. Perhaps you could put some information about the these aspects in your next video. Thanks
Hi! Thanks for the comment. Yes, I plan to make a video about figures for papers. In short, you need to produce vector-based figures (say, in SVG or PDF) from the data you have. You can do this in any software or programming language, for example, plotting packages in Python, Julia, Matlab. Then, you can further edit and improve these figures using vector graphics software (Adobe Illustrator, CorelDRAW, etc.). This is a long creative process where you check if you like the colours, font size and style, alignment of different parts of the figures, and so on. Finally, import the vector-base images into your paper, that is, try to avoid any pixelated figures.
@@chuscience Thanks
Those figures are what happens when you go berserk with Microsoft Visio.
Great and useful content.
Is the literature "radar" you showed at 15:54 original or it's common in your field because this is the first time i see something like it and I'd love to use it in my work, it's way better than a complex table.Also, how did you make it ?
Hi! No, I don't think literature radars are common in my field. It was my initiative to present the literature review in an unusual way. I drew that radar myself in CorelDRAW. You can use other similar software like Inkscape or Adobe Illustrator. Basically, you create a circle, split it into parts (research directions) and manually place papers that you want to mention.
Will be great if u do new video about the source code . For my failed im doing my PhD with communication and information engineering and my research about Mobile edge computing enabled UAV . And how can i get the source code and how basically can code the paper and if i planned to code the paper by myself will led to the results the author shows! I mean optimization problem and mostly using convex optimization . I’m struggling with this part
Hi Hussein! Thanks for the comments. Do you mean developing a code to get results for a paper? I am not sure what coding you are dealing with. If speaking about optimization (say, coding mathematical optimization models in Python or Julia), there are many videos/tutorials on RUclips. Did you see optimization webinars by Gurobi? They have some very basic and clear explanations. Also, this article is useful: www.gurobi.com/resources/mixed-integer-programming-mip-a-primer-on-the-basics/
Minkowski addition to studying the capabilities of Virtual Power Plants is business as usual today, by the way, people who are not in the field still have problems catching this concept.
Thank you
Should the same be done for conference papers or are they more lenient? Thanks for the video!
Hi! Yes, conference papers tend to be more lenient. This is especially true for student conferences (where the majority of submissions come from MSc and PhD students). At such conferences, they decline papers only if they are complete nonsense. However, there are some tough conferences where they expect each paper to be novel and professional. I know a few in my field (e.g., Power Systems Computation Conference) and have heard that the ML community has conferences with rigorous requirements. It is better to check the requirements and acceptance statistics in advance. Good luck!
@@chuscience Thank you so much, this helped a lot!
What platform do you use to create figures please?
Hi! The short answer is: I am using Plots.jl package for Julia programming language; I then save figures in SVG or PDF formats and fine-tune them using CorelDRAW software.
I plan to publish a video tutorial about that (how I create figures) in a month or two. So stay tuned!
Andrey
@@chuscience Will be glad to watch it!! Thanks for your videos ! Really helpful