Hello World - Machine Learning Recipes #1

[@codebasics]

Awesome video, thanks for the insightful conversation over a lunch yesterday. It was awesome meeting with you in person :) Keep up the good work Josh.

[@vi-vf5fv]

For Python3, please note the video does not meet the new syntax. Also note that the video shows 150,0 in the print. It should be 160, 0. To get it to run use: from sklearn import tree features = [[140,1],[130,1],[150,1],[170,0]] labels = [0,0,1,1] clf = tree.DecisionTreeClassifier() clf = clf.fit(features, labels) print(clf.predict([[160,0]]))

[@tjarkraabe5123]

Thank you a lot!

[@phleugh]

I think your [150,1] should read [150,0].

[@gnostict0n1]

Thanks Matthew!

[@manojguha2046]

Thank you very much

[@Paretozen]

When I run the code, my IDE (PyCharm) infers from clf.fit() an object, giving an error on code completion. This made me think my code was wrong. Is there a way to cast or explicitly define a type? I'm fine with no intellisense, just don't like seeing errors :)

[@dmcdeezy]

I have never commented or subscribed to a channel on KZbin. This is how excited I am about this series.

[@amomasi9909]

It all goes downhill from here, mate. Soon you'll be in political debates. Good luck!

[@igorrocha1845]

@@amomasi9909 kkkkkk

[@quantummath]

Thanks a lot for your articulate presentation.

[@jimmylarsson6523]

This is so cool. I just started learning Python, with Machine Learning being the goal, and this is very easy to follow.

[@maxdegreat566]

How's it going

[@reymote2000]

This is the first ML tutorial that spoke to me as a student, not someone who is in a seminar for advanced types. Thank you!

[@NoReason1111]

hey there, it was a nice lecture. Anyways there is an error ... ur code will run if u change the following like from this: print clf.predict([[160, 0]]) to this: print (clf.predict([[160, 0]])) Because In python 3 the print function must have parenthesis, so print(clf.predict([[150, 0]])) I had this error so I thought this might help u also :)

[@alexli3118]

Thanks this helped! I kept on getting that error

[@TroubleMakery]

It isn’t an error, he used python 2.7

[@flueepwrien6587]

@@TroubleMakery smart boy

[@BrotoBhattacharjee]

@@TroubleMakery exactly .. paranthesis got mandatory from Python3

[@gabriel9668]

He installed py 2, not 3

[@rudrakshya1]

I have completed Coursera, udemy machine learning A-Z. Udacity course. But this is a Best getting started I have ever seen for a developer.

[@giuliofagioli]

Awesome, can't wait for next episode!

[@JBGordon]

+Giulio Fagioli Thanks! I'm glad these are helpful. Working on more soon.

[@kuaranir2440]

@@JBGordon Finally, due to you, i have dived into ML :) Thanks)

[@jinxblaze]

This is the perfect way to teach something, explain the history, explain its uses , explain how it originated, explain how people came up with this etc.

[@leep3488]

Awesome! I didnt know machine learning was at the point that libraries could abstract the complexity this much. Finally I can wet my feet in this! Looking forward to the new videos!

[@PeymanEmamian]

This was much better than some random guy (although smart) sitting in front of camera for 500 minutes and trying to explain stuff. I like it. keep it coming.

[@Wintran]

Great video! I love that you show machine learning with tabular data instead of going straight into image recognition like so many other guides. Personally, I'm more interested in statistical data analysis rather than image recognition. Would love to see the same data classified through a TensorFlow neural network. Keep up the great work!

[@manveersingh5822]

I am fascinated by the ML. My GOD, I tested with many other types of data and it made the correct predictions 99% of the time. Thanks for an amazing concise video. GOOGLE is best in everything.

[@vishalsheth1888]

Awesome video! Would be more helpful if the links are in the description below.

[@hrushikeshjoshi8116]

Exactly..

[@donquixoterosinante7899]

He's forcing you to google it.. lmao

[@duncancarr7822]

Thank you! I've watched tens of hours on ML in an effort to find something digestible. And I got the example working! I am so happy. I imagine it is expected behaviour - although I don't understand why ... if I print ( clf.predict([[155, 0]]) ), and print ( clf.predict([[155, 1]]) ) repeatedly I get differing results. I get mostly [1][0], but on occasion I get [1][1]. I used 155 as it is dead centre of 140 & 170.

[@gagginaspinnata]

Love this kind of video on ML. Looking forward for new video!

[@JBGordon]

+gagginaspinnata Thanks for the friendly feedback! Appreciated.

[@JohnJohnson-xt7zf]

This is hands-down the easiest most concise explanation of ML I have ever heard.

[@RoyalBengalCub]

Hi Payne, We are researchers in human-computer interaction (HCI) looking for people who have taken an initiative to recently learn Machine Learning on their own, for career, course or curiosity. It seems you are in that place currently. Would you mind telling us here (www.surveymonkey.ca/r/SelfLearning_ML) about your experiences and any difficulties you faced while self-teaching ML and how you overcame them. There is also a chance to win $50 giftcard. You can help this project by taking out 5-10 minutes to participate in our study. For more details, see here: www.surveymonkey.ca/r/SelfLearning_ML Please share this request with your colleagues or friends who fit this description. People from any major/background may participate. The survey will be open until July 20, 2020.

[@LucasLima-yy5qp]

#version: 3.7.2 from sklearn import tree features = [[140, 1], [130, 1], [150, 0], [170, 0]] labels = ['Orange', 'Orange', 'Apple', 'Apple'] clf = tree.DecisionTreeClassifier() clf = clf.fit(features, labels) print clf.predict([[145,1]])

[@manavkalra5279]

thank youu, i'm lazy af 😂

[@bhavikakapadia2462]

Thanks

[@RoylanMartinez]

#version: 4.4.0 from sklearn import tree features = [[140, 1], [130, 1], [150, 0], [170, 0]] labels = ['Orange', 'Orange', 'Apple', 'Apple'] clf = tree.DecisionTreeClassifier() clf = clf.fit(features, labels) print(clf.predict([[145,1]]))

[@QkdflQkdfl]

Thank you so much

[@ryank.6033]

I wish I could give you a million likes

[@zebcode]

There are loads of fantastic resources into Machine Learning with AI. This has to be one of the best! Thank you Google and Josh Gordon. I really like the presenting style and the enthusiasm that shines through during this tutorial. I get genuinely excited when I start watching this.

[@TheAutophage]

If anyone has trouble with installing the needed software I highly recommend you to first watch this small video Python Machine Learning Tutorial #1 - Introduction from the youtube channel Tech With Tim and then come back to this video.

[@science_mbg]

Your voice is so smooth that makes it easy to grasp I think. I will continue to watch other episodes thanks

[@mahithbhima2064]

This guy just by standing in a single position could explain way better than what our profs could by roaming in the entire room.

[@homeygdoggydogg]

sounds like you were a terrible student

[@Jooma008]

Cant believe these lessons are free. Thanks google!

[@BijoySKochar]

While I've worked a lot on ML and scikit, I am so sure, this is gonna be a show I'll see till the end! Great work Google :D

[@JBGordon]

+Bijoy Singh Thanks! I'm a big fan of scikit. There's a cool library worth checking out - github.com/tensorflow/skflow - that provides a scikit-like interface to TensorFlow.

[@rubeniaborge4652]

Have you worked with graphviz before? Do you know why it gives this error? ModuleNotFoundError: No module named 'graphviz'?

[@sarathchandran2815]

Having knowledge is common,but having the mind to share it with others is great.

[@AbdulMuneerInd]

This is very helpful. Thank you! I'm checking this playlist everyday for the next entries :)

[@kenreid3106]

I have trolled thru a lot of training videos but you sir are a master. You really can get a point across and simplify the hard to understand. Thank You, please do more!

[@laugernberg4817]

if it can detect sarcasm, ill donate 10$....

[@angryguy5532]

very funny guy

[@ComputerProgrammingUniversity]

0

[@bay8834]

I m confided

[@NachoMan154]

"Sarcasm is "a sharp, bitter, or cutting expression or remark; a bitter gibe or taunt". Sarcasm may employ ambivalence, although sarcasm is not necessarily ironic." en.wikipedia.org/wiki/Sarcasm ;-)

[@foriwill7384]

Hey, sometimes i can't even detect sarcasm

[@thinker_211]

So far so clear, im a professional machine learning programmer now. Thank you!

[@forzagreen]

if you want to copy paste the code: ------------------------ from sklearn import tree features = [[140, 1], [130, 1], [150, 0], [170, 0]] labels = [0, 0, 1, 1] clf = tree.DecisionTreeClassifier() clf = clf.fit(features, labels) print clf.predict([[150, 0]])

[@itsmemasud]

+forzagreen I installed the anaconda but where should I do to write the code? I tried with Spyder but an INVALID SYNTAX message shows up as following: print clf.predict([150, 0]) ^ SyntaxError: invalid syntax Can you help me on this?

[@forzagreen]

I guess it's because you are using Python 3. Try with print( clf.predict([150, 0]) )

[@itsmemasud]

+forzagreen After watching your comment I tried with print( clf.predict([[150, 0]]) ) & yahoo it worked but interestingly I installed Python 2.7.11 |Anaconda 4.0.0 (64-bit). Should I use python 3 ?

[@IgorOrlov]

help print( clf.predict([150, 0]) ) - SyntaxError: invalid character in identifier

[@thiagof9475]

Igor , i guess that is print( clf.predict([[150, 0]]) ) you are using a pair of [ ] and its 2 [[ ]] =)

[@marn200]

This is a very clear explanation, I've more learned here in a few minutes than in over all websearch

[@kentonbanyai7917]

Well thats not fair, you can't compare apples and oranges

[@oofusmcdoofus]

But you can differentiate

[@dakshshah2972]

You aren't comparing apples and oranges. You're comparing images, and then classifying them as either apples or oranges.

[@meaniljangra]

I didn't have any idea of machine learning (ML) but after watching this video I know what ML is. This video is too good.

[@011azr]

Please continue, buddy! I'd love to see more of your easy-to-understand explanations :). Thanks a lot!!!

[@shubhamgorlewar3653]

Good 1 to start with ML. Presenter is clear and confident .

[@nishparadox]

looking forward to more of these.

[@adarshashokplr]

The easiest introduction ever seen in the internet , for ML, Thank you

[@BALLAsTvDotCom]

omg release them all please!!!!

[@joshuaporitz4206]

One of the best videos on Machine Learning I've seen. Can't wait for the next one!

[@MOHSINALI-uw5kt]

print(clf.predict([[150,0]])) after print there should be ( )

[@Solid_Fuel]

well yes. you are correct, but only in cases where the version of python is 3.x The code shown in the video is written in python 2. Therefore, it is written without parentheses.

[@pointless132]

understanding = False if understanding : print ("wat?") else: print ("wat?") ==> wat?

[@ahmadbilal3568]

I have taken some machine learning courses, but Seriously Man the way you have teach them is the Best (Thumbs UP)

[@jinxblaze]

Damn, so much better than my uni teachers all combined !

[@fmartins4861]

hahahha :D

[@theshauryasinghofficial]

where to write this code... i have downloaded anaconda ?

[@akshitkumar9402]

@@theshauryasinghofficial in the ide

[@akshitkumar9402]

@@theshauryasinghofficial use jupyter

[@brianandbrain]

This is a really nice, unbiased, beginner friendly introduction to machine learning!

[@JamesCoyle95]

So you will be covering what that library is doing so it could be implemented ourselves? It makes sense for the first video but I kinda hate tutorials that 'solve' a problem by using a library rather than explaining the theory and how it actually works.

[@AsifMehedi]

+James Coyle It's too early to judge. You may like Andrew Ng's course on Coursera, if you want to understand the concepts.

[@JBGordon]

+James Coyle Absolutely! That's the plan. In the next episode, we'll go into a bit more depth on how decision trees work under the hood. We won't get to building one from scratch for a few episodes, though. For a *great* tutorial, check out one of my favorite books: shop.oreilly.com/product/9780596529321.do

[@Wintran]

+James Coyle I think it depends on your needs. I'm actually of the opposite opinion. I dislike tutorials where they teach you everything at the lowest level from start, with too much theory. I want to get up and running fast, learn from examples, and learn to use the best libraries out there. But I agree with you that I want to have a general understanding of what I'm doing, which could be summarized in a few minutes by a good teacher, or included in well-commented example code. Once I get something up and running, I might want to learn more about the algorithms behind it. However, writing a library from scratch is to me never an option. To build your own library that competes against the best ones out there you would not have much time over to actually use it.

[@JamesCoyle95]

Wintran I find libraries are great when you need to interface with something you have no control over or when the task is so complex yet can only really be done in one way. When that is not the case I prefer programming it myself so I have complete control over what happens, how fast it is, how much memory it uses etc. The main problem with many web related tutorials is they usually jump in with 'look how great this jQuery library is, isn't jQuery the best'. It's useless if I need to code in pure JS or even if I need to convert the concept to another language altogether.

[@HivusLibrevilleThiephyst]

I paid more than $50 for online course that can't explain as clear as this channel did. Good job Google

[@Mjiujtsu]

Do you want to install sklearn and follow what he is doing? here's what I did. i have Python 3.5 and windows-64 initially with no other packages installed, so this is "from scratch" you can find out what version of python you have by finding the folder python got installed to, mine was :"C:\Users\Marco\AppData\Local\Programs\Python\Python35-32" but i might have changed that when i installed i cant remember... anyway sklearn requires scipy, scipy requires numpy and nose 1. download numpy usually you can use something called "pip" from the command line which is some magic for downloading packages like sklearn, however, pip is not compatible with scipy and therefore you need to download a (very) slightly different version from: "www.lfd.uci.edu/~gohlke/pythonlibs/" - a site full of windows installers for packages, just use Ctrl+F to find "scipy-0.18.0-cp35-cp35m-win32.whl", you will see mention of MKL included, -that's the small difference i mentioned- .whl is wheel file. if you use pip for this step you'll get and error :"cannot import name 'NUMPY_MKL'" when trying to use scipy NOTE: the cp 35 means that one is for python 3.5, this is the only version that worked for me even though i thought i'd need the win64 version next to it 2.you need to open the command line terminal using "Win + R" and type in "cmd", then type in python -m pip install TheWheelFileName and hit enter, for me, it installed 3. install nose this, i did by pip and encountered no problems with scipy, so in the terminal again type: python -m pip install nose 4. install scipy do the same as you did for installing numpy+MKL, except find scipy on that website, download that and use python -m pip install TheWheelFileName ...again 5. install sklearn the same as installing nose, just do it from the command line using pip python -m pip install sklearn 6. test everything worked by writing into a python module try: import numpy except ImportError: print("fml") try: import scipy except ImportError: print("fml again") figuring this out took me all day, so i really hope that this helps you, because stack overflow can be a bit of a pain sometimes. i found out how to do the pip stuff from the Python Documentation, which has a great part about "Installing modules" and specifically mentions how scipy is not compatible with pip and gave me the link to that website. Have fun with machine learning, praise be to our future robot over lords

[@synetic707x]

Or you just install anaconda and a python IDE ..

[@teleunivsolutions3194]

kzbin.info/www/bejne/ZnvCnnRjhZyNoMk

[@iamsubhrajit10]

thanks buddy

[@omg_look_behind_you]

just like ma used to make

[@nokuphilasimelane5091]

WOW!!! This is the best explanation of ML ever! Clear and Simple, will watch the rest of the series😃

[@SirajRaval]

Hell yeah! If you guys like machine learning check out my new ML series on my channel.

[@logancarvalho]

which software is he using for importing sklearn?

[@MichaelBuergerArt]

anaconda

[@williamfosterXD]

just came from there, it was great!

[@סרטוניםבעמ]

you have a great channel!

[@SirajRaval]

thanks so much!!

[@Phorsakened]

it seems a great and revolutionary move from google that brings courses like this to all . also the main great feature is the simple explanations.

[@Phoenix-flame98]

Your tutorial was very useful, thank you.

[@siamakshams1923]

The following code helped me understand the concept: from sklearn import tree apple = 1 orange = 0 smooth = 1 bumpy = 0 features = [[140, smooth], [130, smooth], [150, bumpy], [170, bumpy]] labels = [apple, apple, orange, orange] clf = tree.DecisionTreeClassifier() clf = clf.fit(features, labels) if clf.predict([[150, bumpy]]): print ("Apple") else: print ("Orange")

[@DomViking]

More of this please :) Just a side note, I received a deprecation warning while testing the code that read: "Passing 1d arrays as data is deprecated in 0.17".

[@itsmemasud]

+Glyn Jackson I installed the anaconda but where should I do to write the code? I tried with Spyder but an INVALID SYNTAX message shows up as following: print clf.predict([150, 0]) ^ SyntaxError: invalid syntax Can you help me on this?

[@vinexio1731]

+Shoaib Ahmed there must be double parentheses after print in python 3 like "print(clf.predict([[150,0]]))"

[@itsmemasud]

Vine Xio hey its working after adding parenthesis .. thanks (Y)

[@Zockermaniac]

@@itsmemasud print clf.predict([150, 0]) -> print clf.predict([[150, 0]]) for versions 2.x for version 3.x it's print(clf.predict([[150, 0]])) so 2.x = print "text" and 3.x is print("text")

[@Mjiujtsu]

man im so happy; i spent all day installing libraries, given that before today i have only tried to install numpy, and failed, and gave up, today i was more determined, but im finally done and i just ran the code he shows at 5:37 and it worked... to save someone the time of all the googling and bs that i did today ill leave another comment with how i did it

[@KawsarJami]

please help me, how do i import sklearn module? here shows: ImportError: No module named sklearn, note that i am using sublime text and i installed scikit learn also, help plz

[@crestfallen706]

this video helped me out a lot looking forward to more videos. switching from software engineering to ML seemed like daunting task but you simplified it, thank you.

[@jim0_o]

I really want to follow a series like this, but at the moment you connect the code together with a gray-code-cloud I'm supposed to accept does a thing and ignore(a library) it just becomes distracting or it feels worthless like I learned something worthless because its all dependent on that gray-code-cloud.

[@SoDamnMetal]

I agree

[@Molloy07]

This is by far the best intro to machine learning I have come across :)

[@robrobbins]

This tutorial could use a little more information on how to use Anaconda. I found Anaconda Navigator and gave the spyder (Scientific PYthon Development EnviRonment) IDE a try and it worked.

[@podlucka]

lol, thanks for this comment ;)

[@mortenbellmann6972]

Really clear speech, fitting speed, really informative!

[@AngelusMortis1000]

I haven't properly payed attention, but I love your voice.

[@bananacluster]

Outstanding communicative clarity! How rare.

[@김진우-q7e]

What should I use IDE in the window?(or recommendation) and tell me how to import(install) sklearn(scikit-learn) in that IDE(you show me)

[@i.d432]

yes all these videos make it sound like you just wave your hands and everything is installed

[@themerpheus]

So,thanks soo much google folks.I wanted to start machine learning,but I have no idea about that.This episodes gives an idea.

[@SuviTuuliAllan]

This guy is a robot.

[@vijayRT]

+Suvi-Tuuli Allan You mean, he's a "machine" Yeah bad joke. I'll see my way out.

[@assafeldman]

+Suvi-Tuuli Allan this guy is not just any robot, its Data

[@chwayalokmen8203]

I have never commented to a channel on KZbin! Thanks a lot Sir.

[@swapanjain892]

when will the next one come out?

[@tytube3001]

+SWAPAN JAIN when you stop touching yourself

[@nopopolasturber]

+SWAPAN JAIN seeing as the #2 episode is up, I admire your self-control and willpower.

[@PrgramedtoMaster]

+Nopopo Lasturber seeing as he wrote that a week ago, I admire your power of observation.

[@theshauryasinghofficial]

where to write this code... i have downloaded anaconda ?

[@howardyang2517]

Not sure if you're still looking after 5 months XD But type it into your text editor and run on your terminal. Go here: learnpythonthehardway.org/book/ex0.html for a quick guide to the difference between terminals and editors

[@rohitmotdhare5688]

I am using PyCharm as Python IDE and sklearn was not preinstalled in it . So to install the sklearn package , File > Settings > Project > Project Interpreter > '+' option > Search sklearn > Install package. Hope this helps, Cheers!

[@frankhubeny958]

I get an "invalid syntax" error at the command "print clf.predict([[150,0]]). The other parts did not generate an error. I am starting with a fresh installation of Anaconda3.

[@李勇-d5k]

thank you, i meet the same problem, after flowing your tips ,fix it.

[@aswinckr]

Pasting response from @forzagreen below Try with print( clf.predict([150, 0]) )

[@IgorOrlov]

help print( clf.predict([150, 0]) ) - SyntaxError: invalid character in identifier

[@aswinckr]

which python are you using? docs.python.org/3.0/whatsnew/3.0.html#print-is-a-function

[@IgorOrlov]

I resolve this problem, thanks :)

[@aniketroy03]

I liked the logic of detecting the more no of orange pixels to differentiate between the two

[@tappiera]

can u do some videos on natural language processing as well? shouldn't be that hard with python i think. Thanks :)

[@RahulSharma2501]

+Enrico Aquilina Natural language processing in Python starts with the NLTK package. Please get an introductory book on Natural language processing and then NLTK cookbook!

[@saiednesbat1870]

Thank you Josh! LOVED this! I'm having an epiphany over this as it'll greatly improve the mobile input methodology I've worked on. Looking forward to the future ones.

[@gneves-w7r]

how in the world do i get to import sklearn.... What should i open... Please...

[@rishikamalik4752]

download anaconda

[@globalwebsmith]

Super easy explanation to clear basics of ML. Expecting more in-depth videos

[@BlindSqrlActual]

"A good feature makes it easy to discriminate . . ." Yeah!

[@mehmetedex]

Woaw this is magical man Im in love with this stuff. plus the extraordinarily simple and clean tutorial ALL THUMBS ARE UP FOR YOU MAN

[@dashankanadeeshan8097]

This is great for a beginner.. Thanks (Y)

[@serge305]

This is exactly what I was looking and has helped me immensely. Keep doing great work!

[@r.d.machinery3749]

Code this up and you can play "What's my fruit?" on your computer.

[@Dr-Brown]

This is brilliant. Made the 6-liner and messed about with it, it's nice to expand it with constants and a dictionary for readability. Now I want to know what else I can do; can't wait for the next episode!

[@itsmeturb0]

reminds me of "Hotdog-not hotdog" 😂😂

[@villajinbernadette2800]

Is this Jin Yang's (idk if my spelling is correct) mobile app on the Silicon Valley series?

[@tejaskulkarni5386]

it was powerful enough for oculus tho :O

[@com48]

Really great work Josh, best introduction to Machine learning I've seen.

[@marcoscal98]

don't recognise "import sklearn" I have installed Anaconda

[@lorenzovarsagod1775]

same

[@ipechman]

Marco Scale did you figured it out?

[@DanishMuneer]

use conda install scikit-learn to install it if uu already got conda

[@roberthelm41]

dank. I've been waiting for something like this for a while

[@3dscan674]

What is the python code editor used in this video?

[@gauravb8]

I believe it's Sublime Text.

[@mayurpuri7786]

IDLE

[@sanghashankar6013]

Atom editor

[@prasanth4997]

Sublime

[@johnofardeal]

awesome! looking forward to see the next episodes!

[@Dominoes0]

I think this vid's a great intro to ML concepts and classification, but the comparison with non-ML techniques is bit misleading - It opens with an example of how difficult it is to classify an image of a apple vs an orange using hand-picked features like color, and described how easy it is to confuse the algorithm with non-standard input. Then it shows a ML example that does the same thing! Except for analyzing number of orange pixels, it analyzes bumpiness. Rather than demonstrating that you need ML to classify this example, as originally stated, it shows that classifying is a lot easier with hand-entered features than parsing an image. Look at it this way - using the number of orange pixels as as a feature might make sense. The real thing that makes the first presentation so difficult is the image parsing - something ML excels at. A more fair comparison might be using pixels as features.

[@JBGordon]

Thanks! Yeah, I agree we could have worked on that example a bit. Hoping to dive into image classification in episode 6 (at a high level), so we can see apples vs. oranges in action. Actually, there's a great code lab folks can try to experiment with that now, called TensorFlow for Poets: codelabs.developers.google.com/codelabs/tensorflow-for-poets/index.html

[@sid007ashish]

You can't expect a beginner to understand it all. At the same time I feel it was right for him not to "throw" all those things on to beginners. Let people get started first, they can learn things on the go. He reduced a huge learning curve I feel.

[@geoffthemusical]

Love the pace of these videos!

[@itsmemasud]

HELP !!! I installed the anaconda but where should I do to write the code? I tried with Spyder but an INVALID SYNTAX message shows up as following: print clf.predict([150, 0]) ^ SyntaxError: invalid syntax Anyone can help?? Thanks :)

[@markankaro1850]

Same error here

[@ryanmatsua33]

If you're using Python 3+, make sure touse print(clf.predict([[150,0]]) or it won't work. Python 3 needs ()s for printing.

[@a-scohen5169]

thanks man

[@gfj555]

Thanks for the great beginners video on machine learning. I knew nothing about machine learning, and even I was able to follow it. Looking forward to the next lesson.

[@rainvdecastro]

Why do i get a Deprecation Warning when i run the code?

[@royalcrew3878]

You have to write it like this "clf.predict([[150, 0]])". Without the brackets it will interpret the input as a "wrong dimensioned" array.

[@rpgtogether7022]

@@royalcrew3878 ITS NOT WHAT HE'S ASKIN' THE PROGRAM RUNS FINE BUT WITH A WARNING

[@amomasi9909]

You're probably using deprecated libraries (i.e. libraries that are no longer being used). So we need to find new way to reference.

[@ALPHADOG1900]

I have never seen a concept so hard to grasp made easy to understand in this videos

[@reephamadness]

"How to get started with machine learning from scratch. To do that, we'll start with two open source libraries." Let me just import these two libraries that each have thousands of lines of code already written. That's like saying I'm going to bake a cake from scratch and going to the store and buying cake mix. Aside from that, good video.

[@realname984]

+reephamadness Is it really? Every time you write some code there is underlying code that gets executed. Example: You make an array. Pretty simple right? Now look at all the underlying code that is necessary for running that code. Do you have to understand it all to use an array? I agree you can profit from learning all the underlying behaviour, but in concepts like neural networks which require a lot of math I think its best for the regular developer to learn to use already existing libraries.

[@reephamadness]

I'm just saying his usage of "from scratch" is improper in this case. From scratch means "from the very beginning, especially without utilizing or relying on any previous work for assistance." That's taken directly from searching "from scratch" in Google's search engine, aka the company that made this video.

[@realname984]

+reephamadness I see what you mean. My definition of "from scratch" is just a bit different.

[@JBGordon]

I see where you're coming from. To write your own algorithm without using any libraries, try episode #5 kzbin.info/www/bejne/d6DIdnuni7iBhdk

[@BUFFDOENUT950]

Well yeah, but you gotta learn from the externals first, then look into making your own machine learning algorithm. It would be sloppy to learn how to make your machine learning AI without first understanding the basics. It is a good start to learn how to call functions from libraries.

[@abhishekwaghmare1543]

If any of the real world teacher could at least remotely tech like you did we could reach next galaxy with many good brains...loved your job

[@simearsov]

Wow, google developers that use a mac book. How do I take your video seriously after this?

[@SgtSayWhat]

Google use macbooks because their Data centers run on linux and the Mac kernal is based on Unix. This "Macs are only for people that know nothing about computers" argument comes from people who know nothing about computers.

[@Anisometry]

Mac has some pretty unique and useful programs

[@mohammedlargo9503]

You have got my full respect you explained it so good that a 15 year old german understood t a piece of this topic

[@EthanArrowood]

Didn't even realize this was in python 2 until the last line xD Darn it!

[@Kev931]

+Ethan Arrowood not to worry scikit-learn.org/stable/index.html supports python >3.3

[@dkarlovi]

+Ethan Arrowood yeah, never really used Python and just found out I'm not on the same version as the video. :)

[@pavelbazin8734]

Thank you Josh and a team behind the scenes!

[@MarkDavey]

Could you please throw this up on github

[@akshaymathur2225]

This is good.. Been searching for this.. Great Job.