Mike is definitely my favorite from Computerphile. As a college computer science student, these videos are amazing for me. I definitely think that this channel should host a niche playlist of more involved programming videos for those of us that are aspiring programmers! Thanks for all of this great content.
Beginning my thesis on Learning Analytics. "LA is the measurement, collection, analysis, and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs" - Siemens, G., & Gašević, D. (2012). Special issue on learning and knowledge analytics. Educational Technology & Society, 15(3), 1-163. Your videos are very helpfiul :) Thanks a lot
You have a talent for teaching and technology. I’m currently interviewing for my first analysis position so this is very reassuring information. Thank you for sharing.
The instance I hear 'data mining' in is usually finding data where it is not easily accessible. For example, you can data mine the console log of a video game to figure out how the game works and whatnot.
When you realize that it's a 10 part series, but it uses R *A small price to pay for salvation* When you see that R indexes from 1 *Reality is often disappointing*
@@sherwinparvizian2414 Have never worked with a programming language that does not start the enumeration at 0. Once more you see that R was developed by statisticians
@Matthias Julia also indexes from 1 , there are other languages out there which index from 1 too , R is built up on FORTRAN and FORTRAN also indexes arrays from 1
This is the series I always waited from Mike Pound! Thank you! But I have a note, at 13:16, high dimensional data (samples with large number of variables) is not necesserily a big data, it's just a very high dimensional data.
@@Wolves2314 But R can suffer from really bad performance; often it is even much worse than CPython. So in that sense I would just use Python or C++. There are some great libraries out there for statistical analysis. Or use Julia which serves the same purpose as R but is much faster and also has awsome tools for plotting.
Are you ready for that? Tell me, truly, are you ready for an algorithm to predict what you need? Imagine the data they need in order to do that. Smart houses coming to a neighborhood near you...
I always thought that Data Mining was acquiring raw data from 3rd parties, like Twitter, Facebook and the like, using their API. For example I once knew of someone trying to predict earthquakes in Japan with Twitter (I don't know how successful this attempt was), but I makes sense to me that this is what Data Mining is -in opposition to working with data you or your company collect, or the cleaner data that other organizations publish.
13:00 I wouldn't agree with your definition of data mining. In a company setting, it's hard to get your hands on a lot of data to start preprocessing, analysing etc, that's when data mining is required: you ask around each department what data they have lying around on hard drives, collect that or spam an API to get more data. So that's prior to preprocessing (because you need the raw material first).
I think most of us make that mistake when we first hear the term "data mining". 😁Once we learn what data mining actually is, then the term makes sense: we "mine" the data we already have to discover valuable information in it. The process of getting data from other departments or other sources is called, prosaically enough, data gathering or data collection. 😊
@@RUclipsr111-p2x yes but they still get to know all about you while you don't know nothing about people that analyse you, your privacy might be protected from other civilians but not from companies that crunch your data
I suppose data sourcing would imply categorizing the actual data you're scraping, right? So in a sense you're pre-processing the data that has yet to be sourced, which of course in turn you're able to (pre-)process further for analysis. Come to think of it, using 'mining' as a definition for it would imply you're digging for gold in land that you might or might not own, hence the ambiguity and why Mike describes it as a buzz word.
11:20 "...and then you end up having 10 saws...don't know how to use any of the saws, but you know, the retailer's job is done..." lol I'm excited about learning how to analyse data using tools like R; but capitalism sure is depressing.
If the first part of the actual series is episode [0], then what number do we give the introduction without some people interpreting it as the last episode in the list?
I thought it was just me until I checked another audio source: sound seems to be «deaf», i.e. lacking a fair amount of treble. I seem to encounter this with more and more youtube videos/channels although I haven't done exhaustive nor detailed analysis, just an observation. Anyway it feels a bit uncomfortable to listen to. Is it also perceived that way by anyone else?
This is my best answer, someone more knowledgable please feel free to correct me if I'm wrong. read.csv() automatically returns a data frame, which is very similar to a matrix, but has additional properties and operations that can be performed on it. If you pass a data frame with higher dimensions, then the function will automatically treat each element in it as 2D unless told otherwise. Judging by how difficult it is to figure out just how to assign higher dimensions to data frames, I'm going to guess that most csv files are 2D and that this isn't a common problem. I'm not sure what you mean by header file, unless you're coming from C++ where header files are a sort of prototype for classes you'd like to implement. C++ is a compiled language, and these header files are preprocessed by the compiler (all lines are gotten from the files) and IIRC, they are not included in the compiled object files. The compiler refers to these files for class and function definitions, so it can know how the class or function is supposed to be structured. It's considered a 'strongly typed' language, where arguments given to a function or class are checked for validity and rejected if invalid according to the definition. R is in contrast to this, because it's an interpreted language and is much more relaxed about type checking. In such languages, types are automatically assumed unless told otherwise. For example, you may need to frequently convert between "factor" and "character" data types in R, because characters are often assumed to be factors (at least, in RStudio they are). Beyond this, you can leave arguments out or explicitly state which argument you're passing to it by name, making the number and order of arguments irrelevant, eliminating the purpose of a header entirely. I hope this helped answer your question! :)
i think 'you may also like' works horribly because its data related to the number of people who bought the item not on data about what they see as missing in their lives. its not peer pressure to have a webpage say 'people also bought" but companies should pay each other to advertise based on tht missing concept
"Big data", "Machine Learning", "Data mining", "AI": Good old computer science and statistics rebranded so Silicon Valley can market old tech as new tech. And investors are buying like Apple fans.
Google: let's sponsor computerphile and Tom Scott to make long playlists that you have to watch, and make people spend more hours on RUclips than 24/day.
No. With python, you get everything R can do with pandas, matplotlib, and numpy, but faster, and with ML with tensorflow/keras. And it's an actual programming language, you can do some actual process with your data then.
Check out the full Data Analysis Learning Playlist: ruclips.net/p/PLzH6n4zXuckpfMu_4Ff8E7Z1behQks5ba
@Z3U5 ]
Mike is definitely my favorite from Computerphile. As a college computer science student, these videos are amazing for me. I definitely think that this channel should host a niche playlist of more involved programming videos for those of us that are aspiring programmers! Thanks for all of this great content.
He is my second favourite. (My favourite is Professor Brailsford )
Big up mike, he's my favourite
agree
10 part series with mike pound yes plz
i love this guy, the way he explains things so clearly and his voice
With open source datasets
>computerphile
>Starting counting from 1
Yeah right
@@JesusisAlive_33 his accent is cool too
I love it when someone who is actually a Dr explains complex topics to me in an accent I can understand.
what accent did you use to hear?
@@hamedal-khateeb7360 Probably indian.
When RUclips recommends you a useful data analysis video, to help you understand, how you always get recommended the other stuff.
Numberphile, NetFlix style: whole season marathon! :) Love it!
(hehe... I meant Computerphile!)
Beginning my thesis on Learning Analytics. "LA is the measurement, collection, analysis, and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs" - Siemens, G., & Gašević, D. (2012). Special issue on learning and knowledge analytics. Educational Technology & Society, 15(3), 1-163. Your videos are very helpfiul :) Thanks a lot
yay two series in one day extending over 2 1/2 hours each
how wonderful
Love for counting the episodes from 0. It helps me a lot with the terms as well. Thank you.
You have a talent for teaching and technology. I’m currently interviewing for my first analysis position so this is very reassuring information. Thank you for sharing.
The instance I hear 'data mining' in is usually finding data where it is not easily accessible. For example, you can data mine the console log of a video game to figure out how the game works and whatnot.
Using R for the explanations, but using a 0 based index for the series titles ... That's not going to be confusing at all 😆
Nope, not confusing at all. 0 ist the first index of EVERY enumeration, makes sense for all of us ;)
@@caughtbynothing R is not zero indexed
@@caughtbynothing really?
@@RIURIU4 Literally every computer language aside from R, Matlab and Fortran are zero-indexed.
@@MaximilianBerkmann Lua
Thank you so much for this amazing playlist. As a CS student, this was very helpful.
I absolutely love the product descriptions for the saws.
Is computerphile transitioning into being a tutorial channel? I like it.
When you realize that it's a 10 part series, but it uses R
*A small price to pay for salvation*
When you see that R indexes from 1
*Reality is often disappointing*
@@krakensoup7439 Mostly convention. A lot of languages use 0-based indexing.
@@sherwinparvizian2414 Have never worked with a programming language that does not start the enumeration at 0. Once more you see that R was developed by statisticians
@Matthias Julia also indexes from 1 , there are other languages out there which index from 1 too , R is built up on FORTRAN and FORTRAN also indexes arrays from 1
ARE YOU SERIOUS, I CANNOT WATCH THIS AND TOM SCOTT AT THE SAME TIME
Same
$DEITY DAMN IT, GOOGLE. I CAN ONLY PAY ATTENTION TO ONE THING AT A TIME.
tom scott?
yup, same
same dude, same
This is the series I always waited from Mike Pound! Thank you!
But I have a note, at 13:16, high dimensional data (samples with large number of variables) is not necesserily a big data, it's just a very high dimensional data.
Was looking forward for this playlist :`). Thanks Mike and Computerphile.
Feel bad for all the people that skipped this video and went to Part 1. lol
R index starts at 1 lol
Just Being Socially Awkward no it's not
@Just Being Socially Awkward you don’t know what you’re talking about and you should feel bad
@@Wolves2314 But R can suffer from really bad performance; often it is even much worse than CPython. So in that sense I would just use Python or C++. There are some great libraries out there for statistical analysis. Or use Julia which serves the same purpose as R but is much faster and also has awsome tools for plotting.
@@nonaviandino8387 True and it shouldn't be the case. But aye, it was designed for statisticians who don't count from 0.
You have no idea how much i need this, Thanks!
I see Dr. Mike, I click.
Doing this in R instead of Python probably won't age well, but I'll still watch it because of Dr. Pound.
Lol the text on the DIY website is really funny! The chainsaw even had The Lumberjack Song xD
An intelligent model is the one that would predict what I NEED; not what is similar to what I looked at in the past.
Are you ready for that? Tell me, truly, are you ready for an algorithm to predict what you need? Imagine the data they need in order to do that. Smart houses coming to a neighborhood near you...
just had a course "introduction into data analysis in R" at uni a few weeks ago. This is a nice revisit :D
Birkbeck, UOL?
And then another series with Robert Miles
Beautifully explained
This tutorial might not be in Python but Python appears at 11:23 with the lumberjack song. Anyone else notice this?
I'm a lumberjack and I'm OK, I sleep all night and I work all day...
Can you guys do a machine learning series as well? This is really amazing!
Dr Mike is Bear Grills of the IT 👍
I always thought that Data Mining was acquiring raw data from 3rd parties, like Twitter, Facebook and the like, using their API. For example I once knew of someone trying to predict earthquakes in Japan with Twitter (I don't know how successful this attempt was), but I makes sense to me that this is what Data Mining is -in opposition to working with data you or your company collect, or the cleaner data that other organizations publish.
Amazing series.
Man I Can't miss any of these videos, ringing those bells :D
A+ product descriptions for those saws
god bless mike pound 🙏🏾
love this guy (i could listen to his voice all day 🤣)
13:00 I wouldn't agree with your definition of data mining. In a company setting, it's hard to get your hands on a lot of data to start preprocessing, analysing etc, that's when data mining is required: you ask around each department what data they have lying around on hard drives, collect that or spam an API to get more data. So that's prior to preprocessing (because you need the raw material first).
I think most of us make that mistake when we first hear the term "data mining". 😁Once we learn what data mining actually is, then the term makes sense: we "mine" the data we already have to discover valuable information in it.
The process of getting data from other departments or other sources is called, prosaically enough, data gathering or data collection. 😊
I will binge watch this
To think this and Tom's playlist come out at the same time :D
We were using "data mining" in place of scraping in the early 2000's. That's likely where that got fuzzed into analyzing and finding use for data.
Data Analysis is important but... Is it possible to have some talk about the boundaries of data analysis regards to personal privacy?
your personal privacy is protected. the names and IPs would be data that would not be used, just the purchases.
@@RUclipsr111-p2x yes but they still get to know all about you while you don't know nothing about people that analyse you, your privacy might be protected from other civilians but not from companies that crunch your data
10 part series with Dr Mike, sign me up
I always knew "Data Mining" to be about sourcing data (e.g. web scraping), not pre-processing your existing data.
I suppose data sourcing would imply categorizing the actual data you're scraping, right? So in a sense you're pre-processing the data that has yet to be sourced, which of course in turn you're able to (pre-)process further for analysis. Come to think of it, using 'mining' as a definition for it would imply you're digging for gold in land that you might or might not own, hence the ambiguity and why Mike describes it as a buzz word.
>starts series at episode zero
>uses program for series which starts arrays at 1
Jasus... this is a great idea for a series, but even an introduction is going to take a year of videos lol. Good luck.
11:20 "...and then you end up having 10 saws...don't know how to use any of the saws, but you know, the retailer's job is done..."
lol I'm excited about learning how to analyse data using tools like R; but capitalism sure is depressing.
Peter Parker:like
Datacamp comment
If the first part of the actual series is episode [0], then what number do we give the introduction without some people interpreting it as the last episode in the list?
i
❤️ loved this!!!
BRILLIANT. THANK YOU!!!
Buzzword soup: "data mining big data with AI-based cloud computing"
these videos should REALLY have subtitles
There will be subtitles for this series? On what it depends if YT will generated it. Automatically generated are fine.
What program did you use to create Browserisor(fake browser)? Thanks
How can I get him to teach me more about computers? Is there a place we can go so that he can specifically teach us or pay for his classes?
I would let Dr. Mike POUND, take me to POUND town any day.
I thought it was just me until I checked another audio source: sound seems to be «deaf», i.e. lacking a fair amount of treble. I seem to encounter this with more and more youtube videos/channels although I haven't done exhaustive nor detailed analysis, just an observation. Anyway it feels a bit uncomfortable to listen to. Is it also perceived that way by anyone else?
Maybe get your ears checked?
@@wiez543 Sorry, can't hear you.
@@nashaut7635 🙉
Great content. Is there a repository where we can get the data files?
Eyyyy Mike uses a thinkpad! My man!
Standard issue at uni I guess?
Mike, what are you going to do with that saw?
Forgive my noob question but how does it know the dimensions of the matrix without a header file?
This is my best answer, someone more knowledgable please feel free to correct me if I'm wrong.
read.csv() automatically returns a data frame, which is very similar to a matrix, but has additional properties and operations that can be performed on it. If you pass a data frame with higher dimensions, then the function will automatically treat each element in it as 2D unless told otherwise.
Judging by how difficult it is to figure out just how to assign higher dimensions to data frames, I'm going to guess that most csv files are 2D and that this isn't a common problem.
I'm not sure what you mean by header file, unless you're coming from C++ where header files are a sort of prototype for classes you'd like to implement. C++ is a compiled language, and these header files are preprocessed by the compiler (all lines are gotten from the files) and IIRC, they are not included in the compiled object files. The compiler refers to these files for class and function definitions, so it can know how the class or function is supposed to be structured. It's considered a 'strongly typed' language, where arguments given to a function or class are checked for validity and rejected if invalid according to the definition.
R is in contrast to this, because it's an interpreted language and is much more relaxed about type checking. In such languages, types are automatically assumed unless told otherwise. For example, you may need to frequently convert between "factor" and "character" data types in R, because characters are often assumed to be factors (at least, in RStudio they are). Beyond this, you can leave arguments out or explicitly state which argument you're passing to it by name, making the number and order of arguments irrelevant, eliminating the purpose of a header entirely.
I hope this helped answer your question! :)
4:38 - vectrices?!
Why all the professors or scientists in the videos of CF use the same kind of sheet? Are you recycling or something?
No mention of sas?
i think 'you may also like' works horribly because its data related to the number of people who bought the item not on data about what they see as missing in their lives. its not peer pressure to have a webpage say 'people also bought" but companies should pay each other to advertise based on tht missing concept
In this series, we will explain how to get people to buy things they'll never use for profit.
What about FORTRAN?
What does this guy not know about 🙌
Great teacher
Somehow I feel like RUclips wants me to become a data scientist.
I'm Mike, and you've just been Pounded
Why doesn't this comment have more likes?!
Of course you start counting from 0
Challenge: take a shot Everytime he says Data
"We as a species produce a lot of data."
ML|AI == Rectangle|Squares
I usually say data mining when I can't think of the word parsing.
please make more on AI and ML and data science
This is beautifull 🤩
The Egss with ai on it are secretly brilliant in germany.
In german 'e'i means egg and it is pronunced just like 'I'
"Big data", "Machine Learning", "Data mining", "AI": Good old computer science and statistics rebranded so Silicon Valley can market old tech as new tech. And investors are buying like Apple fans.
Thumbs up for using R!
How big is big? I don't know.
Which IDE is that?
RStudio
"sponsorship from by Google" - was this piece of English generated by Google's AI?
R right?
Who gave this a dislike?
Why?
Google: let's sponsor computerphile and Tom Scott to make long playlists that you have to watch, and make people spend more hours on RUclips than 24/day.
Lol, an example in R indexed at zero
Excellent, I am python guy, should I learn R ? Is it a good investment of time to learn R ?
No. With python, you get everything R can do with pandas, matplotlib, and numpy, but faster, and with ML with tensorflow/keras. And it's an actual programming language, you can do some actual process with your data then.
I did a Python AI course and realised half way through that what I was being taught was statistical modelling.
AI? Pffffft!
Spider-boy has grown up!
Visualization*
Expecting in python ....
What is data? Am I data?
I have no idea why people would dislike it ?
Ok, R does not equal rstudio. rstudio is one of many IDEs for R and you don't have to use it. I kinda got triggered by the list at the beginning.
i like the 3d effect
0:47 you mean pics from vacations? 😁
Here's a tip: Learn Splunk, get certified, get some experience and earn freaking boat load of cash. Trust me ;)