Intro to Machine Learning (ML Zero to Hero - Part 1)

Machine Learning represents a new paradigm in programming, where instead of programming explicit rules in a language such as Java or C++, you build a system which is trained on data to infer the rules itself. But what does ML actually look like? In part one of Machine Learning Zero to Hero, AI Advocate Laurence Moroney (lmoroney@) walks through a basic Hello World example of building an ML model, introducing ideas which we'll apply in later episodes to a more interesting problem: computer vision.
Try this code out for yourself in the Hello World of Machine Learning → goo.gle/2Zp2ZF3
This video is also subtitled in Chinese, Indonesian, Italian, Japanese, Korean, Portuguese, and Spanish.
Watch more Coding TensorFlow → bit.ly/Coding-...
Subscribe to the TensorFlow channel → bit.ly/2ZtOqA3

[@anakinskywalkerrr]

This man explain what machine learning is in the simplest way I ever heard. Good one, keep it up

[@laurencemoroney655]

Thanks Yoga!

[@d.brilliant5092]

Sooo trueeee!!!!

[@MisterPlatitude]

Laurence, thank you so much for taking the time to put out such concise, intuitive walkthroughs. You manage to make everything going on behind the curtain really accessible and unintimidating!

[@laurencemoroney655]

Thanks! Glad you enjoyed! :)

[@AyshaFilms]

As someone who had just begun self learning programming, this explanation about machine learning is very clear and understandable. Thank you!

[@LaurenceMoroney]

Great to hear Aysha! Thanks! :)

[@ici6308]

Laurence, you're just a genius. I have tried to understand that ML from many tutorials, but it's just from yours I really and simply understand.

[@William_Clinton_Muguai]

In traditional programming, we infer answers after rules act on data, but in ML, we infer rules after answers act on data. Got that really straight.❤️❤️❤️❤️

[@LaurenceMoroney]

Nice!

[@bilboswaggins7629]

Amazing video. Though I do feel the need to say that playing scissors with the thumb out is sketchy and looks like you are trying to straddle the line between scissors and paper.

[@badsanta7356]

It's almost paper like 60 % paper

[@mariomacias4476]

@@badsanta7356 mnb 0:01

[@mariomacias4476]

😮😮😅 nnnjnhj😮😢😅😊😊

[@shashankbarki7029]

Was just waiting for this from Lawrence. I m learning machine learning daily and time to take this to next level. Thanks Lawrence and Google and tensor flow

[@laurencemoroney655]

Thanks, Shashank!

[@M1ndV0yag3r]

@shashank barki would you mind sharing how you are learning ML?

[@brendonprophette8890]

my procrastination has transcended to new levels I am watching this instead of studying for my 2 finals or working on my 4 remaining projects with less than 2 weeks left to finish all of those things lol

[@Intrinsion]

Did you finish?

[@brendonprophette8890]

@@Intrinsion yea, only because my software development professor decided to make the final project optional

[@LaurenceMoroney]

Oops! Sorry about that! :)

[@user-od1kn9mw8i]

can you give me the documentation, and if you would help me you con assist me to make it my final project

[@AshishChauhanYoungy]

Hey there Lawrence. Really good explanation. Thanks for putting together. Just wanted to ask how often these vids will come out?

[@dimonfrekpdimonfrekp3008]

www.coursera.org/instructor/lmoroney

[@ohlssonster]

Once per second

[@laurencemoroney655]

This series is 4 videos, coming out weekly

[@javiersuarez8415]

@@laurencemoroney655 4 is a small number. 😐. When is estimated second season release?

[@LaurenceMoroney]

@@javiersuarez8415 Haha -- I haven't gotten around to filming a second season yet, but as they look like they're going to be popular, I should get moving on that... :)

[@DirkJanUittenbogaard]

Great video Laurence! For me the code you used failed "ValueError: Unrecognized data type: x=[10.0]". After changing the last line (print model predict) to this it worked: print(model.predict(tf.convert_to_tensor([10.0])))

[@hamsterman1571]

giving xs and ys are array but as an input u are using a list '10.0' so its error, u can also try : predict(np.array([10.0])))

[@Themojii]

Subscribed after watching this. Love the way you explain. You explain the concept very clearly and also you add a little bit of the code which gives me a great preparation for the coding application. Keep up the good work Lawrence

[@vishalrana1373]

Brilliant explanation Laurence.

[@laurencemoroney655]

THanks, Vishal

[@DatascienceConcepts]

Nice explanation. I am also building a course on ML in Python (for a University) more from an implementation perspective. This surely helps!

[@albertastillero1085]

Subscribed Sir Laurence! Thanks for the simple yet concise explanation in a short time.

[@jesjames]

You know those videos that you start watching and then get glued to them... :D Well done Laurence, in the first few seconds I wouldn't have bet on watching it

[@laurencemoroney655]

THanks Jes!

[@shubhamdeep1804]

@@laurencemoroney655 k

[@aytunch]

Laurence keep more videos coming:) Was a pleasure watching and learning

[@laurencemoroney655]

Working on it! :)

[@fet1612]

Hello, Laurence Moroney, Astounding presentation. How quickly and how brilliantly you put such a huge task look so simple. I must admire your ability. Keep up the good work, thumbs up here.

[@laurencemoroney655]

Thanks Fet!

[@keithprossickartist]

Thank You for explaining this so clearly and eloquently.

[@ehsankiani542]

You're genius Laurence, for sure! Excellent demonstrations and brilliant examples.

[@laurencemoroney655]

Thanks!

[@mihaelacosinschi]

here's the code needed from the video, if anyone wants to try it out import tensorflow as tf from tensorflow import keras import numpy as np model=keras.Sequential([keras.layers.Dense(units=1, input_shape=[1])]) model.compile(optimizer="sgd", loss="mean_squared_error") xs=np.array([-1.0, 0.0, 1.0, 2.0, 3.0, 4.0], dtype=float) ys=np.array([-3.0, -1.0, 1.0, 3.0, 5.0, 7.0], dtype=float) model.fit(xs, ys, epochs=1100) print(model.predict([10.0])) I am an absolute beginner and wanted to run the code, but could only get errors at first. In case anyone needs this broken down, I added the first 3 lines that are necessary to run the tensorflow and keras libraries, installed previously via terminal.

[@saedsaify9944]

The code is wrong. Not a good sign when the Hello World code from the official channel doesnt work. print(model.predict([10.0])) throws an error, you need to use something like print(model.predict(x=np.array([10.0])))

[@guyshur2688]

Great video. You mention the small error is due to uncertainty due to the low sample size, is it not possible that the model simply descended to a not quite accurate relationship? Granted the cause would still be low sampling but the main question is if the error is explicitly programmed to reflect uncertainty because the input could still be 19 and be labeled uncertain.

[@SK-lm2zs]

this video is soo good❣️ I watched this many times to understand what ML is. I studied Matlab at University, this video is also good for review of ML😊

[@sriti_hikari]

That was a very good explanation, thank you!

[@nathanas64]

This is one of the clearest explanations ever ! Great job!

[@laurencemoroney655]

Thanks!

[@nathanas64]

Laurence Moroney what a service to humanity that google is releasing tensorflow to the public domain. The benefit that will come out of this -and i don’t mean financial - is immeasurable. It’s like IBM releasing the paper on FFTs in the 60s !!

[@AlexeyArtamoshin]

Extremely helpful explanation, thank you very much!

[@johnlao1469]

Im really glad tensorflow by itself is doing tutorial right now. Because i have this research project that implements machine learning and it helps me to learn and understand each lesson about it.

[@laurencemoroney655]

That's great, thanks for letting us know! :)

[@venugopalt6861]

This is awesome mike the best explanations i have ever made on Machine Learning and i got a feel and beauty of nerual network when i heard your class , great job , keep posting like these cheers

[@LaurenceMoroney]

Thanks so much! :)

[@rohanmanchanda5250]

You're welcome kid.

[@rohanmanchanda5250]

@@LaurenceMoroney I didn't notice it was actually you sir. This function cannot recognise polynomials like square equations or cubic. I provided it with xs as 1.0, 2.0... and ys as their square, but it never got any better than a loss of 6.2222, and if I entered 10, it gave me a value of 36.67... ???

[@marcosdearruda77]

Nicely done. Thank you so much for sharing this video with us,

[@joseortiz_io]

I love these tutorials and videos that Tensorflow puts out. Super informative. Thank you Laurence, what a great video! You bet I'll keep watching these series! Have a fantastic day everyone!😁👍

[@laurencemoroney655]

Thanks! :)

[@PareshTadvi-qr1to]

...

[@PareshTadvi-qr1to]

..

[@PareshTadvi-qr1to]

Nj... N

[@dipankarkaushik5285]

You missed 'tf.keras.' in the 1st line. So, model = tf.keras.Sequential([keras.layers.Dense(units=1, input_shape=[1])]) will be the correct code.

[@LaurenceMoroney]

oops!

[@ConsultingjoeOnline]

Great video! THANK YOU. I've been trying to get to this point for a while. Getting everything setup is a hurdle in itself. At least with OSX

[@deborahayeni_]

Thanks. You just spiked my interest in this course

[@autosales3196]

Rinse spin repeat.×3 or X4 to remove one situation. ......I can do this. Thanks for the patience ☺️ God sure made a blessing in you!

[@florianeck6512]

finally some video that makes digging into the topic understandable.

[@LaurenceMoroney]

Thanks!

[@user-sd2cd2vj1f]

Great, more inquisitive on the subject

[@donkeshwarkavyasree8632]

This video's are literally making me feel fascinated to learn ML. You are definately life saver 🙏. Thanks a ton 👍

[@LaurenceMoroney]

Very welcome! :)

[@ademolaorolu5930]

Precise and Concise. Thank you Lawrence!

[@rishishkumarnaik9954]

Hey Lawrence, its really a pleasure to learn from your videos. Waiting for more videos to come and take us deep into AI.

[@laurencemoroney655]

Thanks Rishish! :)

[@kamilpadula7152]

this open for me new world

[@khadijahalsmiere3718]

Wow , I’ve been waiting for such an opportunity to learn machine learning from an expert . Thank so much and keep it up , we need it for our big project GOD’s willing .

[@laurencemoroney655]

I hope it works out :)

[@khadijahalsmiere3718]

Thanks a lot

[@logicfacts9964]

I didn't understood what exactly is input shape and why it is 1? Because is accepts our input array by only 1 value at the time or there is other reason? Also I can't understand how and why NN with just 1 neuron produces 18.99 instead 19 because 1 neuron means that it can predict only exact value and any deviation is inposible?

[@laurencemoroney655]

Input shape is 1, because we just want to predict the result for 1 value input (i.e. 10). Neuron won't get *exact* value because it deals in probabilities, not certainties, so the prediction is a very high probability that the answer is 19, but when evaluating that as a number you get something close to 19

[@koushikdatta5130]

Loved the intro. Waiting for the next video.i was searching for such tutorial for long time, finally got one. Thanks tensor flow.

[@laurencemoroney655]

Welcome! Thanks for watching!

[@fahemhamou6170]

Thank you very much

[@jng711b]

Great explanation. I am taking Lawrence's courses in ML / Tensorflow. Very useful. Thanks so much!

[@LaurenceMoroney]

Thanks John!

[@the_bitcoin_guy]

I feel like Neo : "I know kung fu 🥋! " . That was so concise !!! Thank you very much ...

[@LaurenceMoroney]

haha! Thanks :)

[@alex9046]

Laurence will do that to you lol, amazing teacher

[@hobypd]

I have a question on something I don't understand: Dr. Moroney said that prediction is not perfect because the computer is trained for 6 values that form a straight line, but outside those 6 may be not straight (although it is highly probable that they are straight). I don't get this point: since it is a NN with only one neuron, so it has to be a straight line the prediction (it should be like a linear regression). Am I correct? Or did I interpret something wrong?

[@aradaizlebeni1]

very good and simple lecture. thank you.

[@Pa-ow1nj]

please more of that its so good explained

[@laurencemoroney655]

Working on it! :)

[@jfl1014]

Very good teacher thank you

[@ashwinsenthilvel4976]

Hi Lawrence, I am trying to implement same code with two inputs X1 and x2. I am finding difficulty in 1) how to specify x value like how the matrix of the two input should be. 2)what must be the input shape specified here. Could you please help with this.

[@moacirfranciso9645]

Hinos

[@MikesTechCorner]

In the math example we get a NAN when typing in other x values in the array like 100. Do you know why? from __future__ import absolute_import, division, print_function, unicode_literals import tensorflow as tf import numpy as np x = [-1.0, 2.0,4.0,6.0,7.0, 100.0] y = [] x_test = [10] for i in x: y.append(i*2 +5) model = tf.keras.models.Sequential([ tf.keras.layers.Dense(units=1, input_shape=[1]) ]) optim='sgd' model.compile(optimizer=optim, loss='mean_squared_error') xs = np.array(x, dtype=float) ys = np.array(y, dtype=float) model.fit(xs, ys, epochs=500) print(model.predict(x_test))

[@LaurenceMoroney]

Data should really be normalized when fed in for training, or the optimizer/loss won't work. We get away with it when we use small values, but that gets exposed at larger values. To do this you should normalize the training/test data and then retrain.

[@theprimordialdude1138]

Very good explanation. Easy to understand. Continue the series

[@LaurenceMoroney]

We are :)

[@gennadyplyushchev1465]

Great and simple video! Thank you!

[@laurencemoroney655]

Welcome! :)

[@zaknikov]

Very good explanation thank you

[@laurencemoroney655]

Welcome! :)

[@quegiangson3786]

It's so amazing explanation. Thanks a lot Lawrenece !

[@laurencemoroney655]

Thank you! :)

[@AbhishekKumar-mq1tt]

thank u for this awesome video

[@laurencemoroney655]

You are welcome! :)

[@AnandBaburajan]

Here's the working code: from tensorflow import keras from keras.models import Sequential from keras.layers import Dense import numpy as np model = keras.Sequential([keras.layers.Dense(units=1, input_shape=[1])]) model.compile(optimizer='sgd', loss='mean_squared_error') x=np.array([-1.0,0.0,1.0,2.0,3.0,4.0], dtype=float) y=np.array([-3.0,-1.0,1.0,3.0,5.0,7.0], dtype=float) model.fit(xs,ys,epochs=500) x1=np.array([10.0], dtype=float) print(model.predict([x1]))

[@nabhannoorish5100]

Thanks for teaching this. You made this very easy

[@thecandel5479]

You are very very good scientist. I thank you very much. I am from Jordan. I study master in computer and networks.

[@laurencemoroney655]

Thanks Raed!

[@r.a.9802]

Gracias, thank you, danke, merci

[@acutu55]

Great Introduction!

[@simonwhitehead2857]

Ok, you show some code that builds and trains the model before making a prediction. I found that on subsequent runs the accuracy increases, I realize that for some applications this can result in 'overfitting'. So once I am happy with the level of accuracy ,how can I apply the trained model without running the training (how/where is the model saved?)? Really love this course my head is working overtime in thinking of ways I want to try and apply it!

[@SamuelLawson]

Running that code, I got an error: `ValueError: Unrecognized data type: x=[10.0] (of type )` -- fixed when I changed the predict() arg to `np.array([10.0])`

[@skewd2528]

Do print(model.predict(np.array([10.0], dtype=float))) instead

[@NicO-cm2xo]

Awesome master teacher Lawrence.. now i need autoML to learn ML

[@laurencemoroney655]

Haha, so do I! :)

[@muhendishanimm]

I really loved the videos then liked all before watching cos i am sure I will watch all :D inshaallah :D Thanks for yoru effort!

[@LaurenceMoroney]

Welcome!

[@BiancaAguglia]

Nice, clear explanations. This series is off to a good start. 😊 Looking forward to seeing more videos.

[@laurencemoroney655]

Thanks Bianca!

[@psaikrishna90]

Wow such a great explanation with a simple example. Thanks.

[@laurencemoroney655]

Thanks! :)

[@abdenourdjennane9309]

Thanks. I like this way of teaching.

[@researchforumonline]

Good explanation, thanks.

[@zhang20244]

So nice , easy to understand , Thanks

[@laurencemoroney655]

Glad you like! :)

[@ericascheffel4527]

Excellent explanation, thank you very much!

[@LaurenceMoroney]

You're welcome, Erica!

[@sunilkb144]

i have typical use case such as to predict the value of y for a given x but the logic for calculating the value of y ( ie in this case y=2x-1) is changing on daily basis. can tensor flow can predict this kind of data

[@laurencemoroney655]

If data changes, model should be retrained

[@naduniranasinghe6593]

super explanation. you are a great teacher

[@LaurenceMoroney]

Thanks Naduni!

[@murmodeus]

I tried to feed the values for "y = x² + 5" but the algorithm failed to predict the result. I guess this setting is only good for "y = mx + n" kind of equations. I wish I'd understand the reason though.

[@MikeKay1978]

I did the same and I guess since there is only one neuron it can only find one weight i.e factor and a bias i.e offset. So you will need more neuron i.e a bigger nn to be able to handle x^2 polynomials.

[@Arnau_0_0]

Really good explanation!

[@codewithluq]

very nice way of teaching

[@LaurenceMoroney]

I'm trying! :)

[@papareddysrinivasulureddy7591]

Sir is it important to study complete process of all machine learning algorithms or it's just enough to know the application of each algorithm . please tell

[@febiri6979]

I really appreciate you Sir

[@laurencemoroney655]

And I you! :)

[@slvrpltd]

stupid question: where do i tell the thing that the relation between x and y is a math problem? how does it know that it has to calculate something and not come up with something else?

[@krishnachauhan2850]

Please make a series on audio data loading n analysis using tensorflow

[@kvenkataraju]

Thanks for explanation, if I feed x and y with more values like xs = 2000.0 and ys = 3999.0 (2x-1), my loss becoming infinite. Why that is happening? Great if you share more details on it.

[@laurencemoroney655]

Data being fed into the neural network should really be normalized. I.e. made to be between 0 and 1. We get away with not doing that with the smaller values here, but if we don't d that, they we'll usually end up breaking the loss function / optimizer and getting infinite loss etc. So if your range of Xs were from 0 to 2000, and your range of training Ys were (2x-1), you should divide all Xs by 2000, divide all Ys by 2000, train on that, and when you do a prediction, divide the input by 2000, and multiply the output by 2000 etc.

[@adityadudeja1126]

on which data losses are calculated because the model has seen train_data and train_labels

[@laurencemoroney655]

It has seen those, and it calculates loss using those. The model will measure its performance against the known values, and calculate which ones it got 'right' and which ones it got 'wrong', with the loss function reporting loss on these.

[@rhenium1877]

Can you make a video about Multiple Linear regression without using any python Libraries.I mean the math behind those libraries.

[@alessandroruggiero8932]

Can i use purhon 3.7 or have i to use 3.6 as before

[@laurencemoroney655]

Sure

[@rajansaharaju1427]

eagerly waiting for part-2

[@laurencemoroney655]

It's alredy out. Part 3 next week.

[@prathameshdinkar2966]

Nice! but there is a bug in the last example of the colab sheet - when u try to stop training if the accuracy becomes more than 90%

[@laurencemoroney655]

I'll check it out

[@hemakumargantepallidataandai]

Awesome presentations skills.

[@piers9186]

Excellent. When does No.2 arrive?!

[@laurencemoroney655]

Weekly

[@angeloabritaa]

Bom tutorial, aguardando continuação. Like from Brazil hu3br

[@LaurenceMoroney]

Thank you! :)

[@chrismorris5241]

Rules + data vs. answers + data. Pretty good.

[@laurencemoroney655]

Thanks!

[@laurencemoroney655]

Thanks!

[@sabaal-jalal3710]

clear explanation thank you so much!

[@LaurenceMoroney]

THanks!

[@abdallahhussein5997]

Perfect explanation

[@LaurenceMoroney]

Thanks!

[@kartikaykhosla1201]

Hey Lawrence just wanted to know is it a weekly series or will just come whenever next is available

[@laurencemoroney655]

Goal is weekly

[@regeleionescu935]

Hi, thank you for the great video! I have been playing around with your example using different sets of numbers. For example, I extended the first set from 1.0 to 12.0 and for the second set I used the days of the month(like 31, 28, 31, 30 etc). With a set of 12 pairs it worked fine. It managed to predict the 13th month as being 31 days. Then I wanted to be smarter and I extended the set of data to 24 pairs, basically repeating one year and I tried to predict the 25th and the 26th month length. The problem is that with 24 pairs of numbers the error keeps growing to infinite and the final result is NaN - I wonder why?

[@laurencemoroney655]

Yeah -- when you start using a lot of numbers the error rate grows, because the dataset isn't normalized. Try normalizing, then learning from the normalized values, and then denormalizing afterwards.

[@regeleionescu935]

@@laurencemoroney655 Thank you very much for your kind answer!

[@kvenkataraju]

@@laurencemoroney655 Hi, you mean normalizing the data? how to do that? is that by adding normlization layer?

[@mm148881]

print(model.predict(np.array([[10.0]])))

[@TudoEstudado]

this is so much difficult how do I landed here? try to only start something but I do not know even where to start. too much info

[@TheCyberCraft]

Lawrence, great job man! thank you so much

[@laurencemoroney655]

Thanks! :)

[@fpgamachine]

Excellent video! Thanks

[@laurencemoroney655]

My pleasure! :)