Understanding AI - Lesson 1 / 15: A Simple Neural Network

PLAYLIST: • Self-driving Car :: Ph...
Embark on an exciting journey into the world of Artificial Intelligence with my new course, "Understanding AI". In this first lesson, we dive into the fundamentals of Neural Networks, breaking away from the conventional Machine Learning context. Try out this unique hands-on experience where you'll manually tweak network parameters to teach a car how to drive within a specially crafted playground.
Whether you're new to AI or seeking a deeper understanding, this course caters to all levels. I have over a decade of machine learning expertise, and will emphasize the importance of revisiting basics in an era of easy-to-use complex models. Homework assignments and a final challenge to race AI cars await you, with live stream events and prizes on the horizon.
The course covers essential topics such as Dijkstra's shortest path algorithm, game mechanics, camera sensor creation, and implementing analog steering through mobile apps and image processing with augmented reality. All implemented in plain JavaScript without external libraries, giving you a comprehensive understanding of the system's inner workings.
AI is more than just Neural Networks - it's a combination of various elements. This course demystifies the distinction between Artificial Intelligence and Machine Learning. While there's no Machine Learning in this course, I do plan to teach that someday as well. If you want to help collect data, go here and race:
🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙🚙
radufromfinland.com/projects/...
🚗THE PLAYGROUND🚗
radufromfinland.com/projects/...
⭐LINKS⭐
Phase 1: • Self-driving Car :: Ph...
Phase 2: • Self-driving Car :: Ph...
Prerequisites: • Math & Code Fundamentals
Machine Learning Course: • Machine Learning Course
3b1b's Neural Network Video: • But what is a neural n...
Another Playground: playground.tensorflow.org
💬DISCORD💬
discord.gg/gJFcF5XVn9
💻CODE💻
1. Phase 1 github.com/gniziemazity/self-...
2. Phase 2 github.com/gniziemazity/virtu...
☕Buy me a Coffee?☕
www.buymeacoffee.com/radum
⚡️Join this Channel⚡️
youtube.com/@radu/join
Excited to get started? Put your neurons into overdrive and join me on this fascinating AI adventure! 🚗💻
#UnderstandingAI #AIPlayground #NeuralNetworks #SelfDrivingCar #AICourse #javascript
⭐TIMESTAMPS⭐
00:00 Introduction
06:34 The Playground
14:40 One Neuron
21:48 Decision Boundary Plot
25:05 Homework 1
29:05 Homework 2
29:13 Clarrifications

[@mandelatrevor2116]

WHAT A COURSE!! This is what the internet was made for. Thank you Radu 👏👏

[@Radu]

Glad you liked it! :-)

[@shaddow11ro]

Right! Imagine if Radu was teaching when i was in school!

[@themightysapien]

The dungeon master is finally here. THE MAIN BOSS.🔥🔥🔥 Thank you for all the work you do behind the scene to put up such mind blowing lessons.

[@Radu]

There was a lot of work in this one, for sure... Thanks for watching :-)

[@JuanVargas-bv5yp]

What a course! this is insane!

[@serveshchaturvedi2034]

You should rename the playlist to building your own Tesla autopilot 🎉😂, simply amazing. Thank you for this

[@Radu]

It's nowhere near as advanced, though :-)

[@edgaremmanuel3197]

i think for the KZbin algorithm you should definitely rename it to build your own Tesla like autopilot

[@Radu]

Haha, I don't care about the views that much :-)

[@wilfredomartel7781]

😊amazing video man.

[@AyorindeAdesugba]

Wonderful Friday - thanks for everything

[@Radu]

Thanks for your support :-)

[@alexandersolano227]

Thank you, Hokage Radu of the IA village. Your contribution is invaluable to the new IA shinobis.

[@Radu]

:-)) thanks

[@airbornesnail]

What are you? Some kind of gift to earthlings? Never seen such a calm, thoughtful, fun, easy to understand coder in whole my life. Great work. I truly hope YT sends money to you in 24m trucks.

[@Radu]

:-)) maybe it's because I'm not doing it for the money

[@lucasdecarvalho654]

Man, you're the best gift that internet gave me! Woow! What a course, dude!

[@Radu]

Thank you :-)

[@AlexBegey]

Radu, you're awesome! Love your videos & your teaching format. Very interesting, useful, and inspiring! Looking forward to the next videos.

[@Radu]

Glad to hear you find them worth it :-)

[@ianchaplin2777]

these courses are amazing. Can't find any series anywhere on youtube nearly as good as yours. Keep up the good Work! Can't wait until the next video!!! :D

[@Radu]

Thank you!

[@kpm25]

Thanks for your hard work Radu I've dilligently gone through Phase 1 and 2 and many of your other tutorials, much appreciated!

[@Radu]

Thanks for watching :-)

[@ryanest3684]

Radu!! I need the next video!! You explain everything so well!

[@Radu]

Coming out on Friday!

[@mohammadsaifulislam8341]

Great tutorial, Thanks Radu for this awesome playlist ❤

[@Radu]

No problem 😊

[@mahmudadamusman8264]

Thank You So Much Radu! Everyone else confuses me, but even before watching this series: I am happy for I know you are The Right Teacher. I am excited, truly.

[@Radu]

Glad to hear you understand :-)

[@andyj8977]

Absolutely brilliant! Thank you so much for this Radu 😎

[@Radu]

More to come :-) stay tuned next week!

[@mathematicsclub961]

I have been waiting for this 😃. Thank you for uploading

[@Radu]

Thanks for waiting :-)

[@__angle]

Thank you so much for your great explaination and humour ! I'm happy to follow

[@Radu]

You're welcome :-)

[@stebansb]

So glad I found your channel, this is just amazing.

[@Radu]

Glad you like it! :-)

[@Ravroid]

Thanks so much for putting this course together, this is tremendously valuable

[@Radu]

Glad you think so! Thanks for watching :-)

[@austinfinell]

I'm so excited to get this far! I'm still working on the first stage of tutorials to really get a good understanding of my foundations, but I wanted to tune in and comment to show my support. Going through and making sure I understand each line of code is how I'm getting started with JavaScript. It's taking a bit longer than I expected, but I think it will be worth it. Thank you so much for putting in all of this work to help us learn!

[@Radu]

I see. If you need a break from JavaScript you can look at the first lessons of Phase 3 with no worries. There won't be any coding in them.

[@fdorsman]

I have been waiting for this phase. Thanks!

[@Radu]

Hope you like it :-)

[@fdorsman]

@@RaduI sure did! Can't wait for the next episodes

[@Radu]

@@fdorsman cool :-)

[@warriorsofxu8394]

Always a joy to learn from your videos😊

[@Radu]

happy to hear :-)

[@shaddow11ro]

As usual... Radu is always baking something extraordinary!

[@Radu]

Hope I don't forget the oven on :-D

[@adilsonbuset738]

Muito obrigado, mestre Radu!😃

[@Radu]

You're welcome! :-)

[@redgood5879]

i cant say how much ive learnt from this course thank you radu !

[@Radu]

Great :-) More to come!

[@Merilix2]

Homework 1: I guess there are as many solutions as 2^ (as many mantissa bits as floating point operations can deal with). Great course by the way :)

[@Radu]

Oh wow :-) cool answer! I guess you could also implement your own way of defining large numbers (like using arrays?) and doing operations on them?

[@Android-17]

Thank you Radu🙌🏻

[@Radu]

You’re welcome 😊

[@user-ey9wf6ty5c]

You are amazing!!! Anothers thanks for sharing your knowledge and specially for your efforts!

[@Radu]

Thanks you for watching :-)

[@eridarael6541]

Thanks very much Radu. I can understand now how neural network work. I have been confused with many other tutorials. It's more easier to understand while implementing it in a project. Thanks very much

[@Radu]

Much more to learn in the next lessons! Thanks for watching :-)

[@merarebbadro9328]

Thank you, you are amazing. I am looking forward to your videos

[@Radu]

Thanks for watching! :-)

[@martinarce8138]

Radu rocks!! Thank you for sharing your passion!

[@Radu]

My pleasure! :-)

[@dconnectzone]

Wow. Thank you for this. Currently learning AI.

[@Radu]

Best of luck!

[@CarlosToscanoOchoa]

JESUS CHRIST 🤯 No libraries?? Absolutely mind-blowing 👏🏼👏🏼

[@Radu]

:-))

[@emilyhiroshi7822]

thank you so much for your amazing work!

[@Radu]

My pleasure! :-)

[@houbill5363]

Great course!!!! Thanks a lot!!

[@Radu]

You are welcome!

[@user-ys9rn7zv7x]

Thank you for the tutorial

[@Radu]

You’re welcome 😊

[@aGj2fiebP3ekso7wQpnd1Lhd]

Great video Dr. Radu. 🎵 Coding with Radu... Coding with Radu... boom

[@Radu]

Haha, thanks :-)

[@seiwaldm]

omg 😱 loved phase two (still have to watch phase one)... but wasn't anticipating that much extra content in phase 3... 3d sensor view? steering with smartphone or webcam? blows my mind 🤯

[@Radu]

Glad to hear you're looking forward to it :-)

[@alexandersolano227]

¡Gracias!

[@Radu]

Thanks!

[@___d3p1]

You are amazing! ❤

[@Radu]

:-)

[@ypohut1673]

Спасибо🙏💕 ты молодец!

[@Radu]

Thanks for watching :-)

[@kingcoder31aug]

Very Very excited

[@Radu]

Cool!

[@stevehsu777]

Hi Radu, the video is so great!

[@Radu]

Thanks a lot! :-)

[@kingothemonkey]

Getting in on day one!

[@Radu]

Good luck :-)

[@ivgadev]

Yes!

[@Radu]

:-)

[@firefox_tn1815]

amazing 🎉 i finally watched it

[@Radu]

Thanks for waiting :-)

[@mauriciosantecchia6794]

Thanks radu, you are a master explaining, I love your courses. I don't know if you have some tutorial about this, but I should love if you do a course about reinforcement learning to train a neural network, maybe taking you previous tutorial about self-driving car... just an idea. Thanks for all your work, I'd like that your videos have more likes, you deserve it!!

[@Radu]

I guess you can try adding reinforcement learning to the self-driving car code from phase 1... Some people consider the genetic algorithm we used as a kind of reinforcement learning... where the fitness function acts as a kind of 'reward'... it does generate multiple instances and proceeds with new generations, though... so, I wouldn't call it reinforcement learning because of that. I would expect one agent doing things and being rewarded / punished appropriately. What is not so obvious is how to give these rewards and punishments, really.

[@mauriciosantecchia6794]

@@RaduAnd while we are on the subject of GA: I tried to make a change in the self-driving car project by having only a small percentage of the population mutate and taking the best members of the population and crossing them with each other. In this way I hoped to reduce the destructive effect that randomness has during mutations and preserve the good characteristics acquired by the fittest members. Short summary of the results of the experiment: it did not work :( I think I have to do more research on how to cross NNs because I tried to mix with precise patterns the weights and biases and that did not work. I also tried other things like averaging and variants of it, but it got worse :P . Evidently I lack more mathematical knowledge about NNs functioning to do this process better. Any suggestions on how to cross? (note: to calculate the fitness I took into account the road covered and the amount of cars overpassed)

[@Radu]

I think that some of the things we play with in this new course can shed some light on why those techniques don't work so well.

[@disrael2101]

amazing radu finally the AI chapter is here, which to me is the most interesting by far!! how often do you think you'll upload new lessons / chapters, once a week or? either way I can't wait for a complete AI real-world applications bootcamp (without libraries) by you , i'm willing to deposit money now for it !!

[@Radu]

New lessons will come out every Friday until mid-May.

[@Barfmat]

Thanks

[@Radu]

Thanks!

[@volodyslove]

You are the best 😁

[@Radu]

Thanks :-)

[@mdsalahuddin46464]

Great 🔥

[@Radu]

Thanks! :-)

[@vijayserpentj7910]

Awesome

[@Radu]

Thanks! :-)

[@SufianDira]

Homework one solution: There are infinitely many. Because there are infinitely many lines that intersects the x-axis at any point.

[@Radu]

Good :-)

[@usama_khalid]

superb

[@Radu]

Thanks 🤗

[@Xoundz]

Love u bro ❣️

[@Radu]

❤

[@edgarsvilums1550]

First 5 minutes. Mind. Blown.

[@Radu]

And then it became boring? :-D

[@edgarsvilums1550]

I think we all know the answer to that. :D Thanks for your work, really!

[@Radu]

You're welcome :-)

[@tomekatomek5694]

I love you for that video 🙃

[@Radu]

:-))

[@MajidFouladpour]

Initially I was confused about the relation between the bias value and the point at which the car stops. Then I realized that although we stop accelerating the car at point y (where the neuron turns off), it will keep going forward to point y' because it has some inertia.

[@Radu]

Yeah! It's a good idea to use the manual override in the playground and see what the car really is like. The neural network will need to consider its capabilities as well. In this case it's not so much about inertia, but about the friction. The friction coefficient is fixed in our implementation from phase 1, but it could be a variable (depending on the outside temperature, road surface, etc). In that case the neural network needs to know those conditions as well. But here we won't have to worry about small things like that, we'll add complexity some other ways. You'll see :-)

[@karolisr]

Yessss

[@Radu]

:-) hope it will be worth it

[@Yrgruuc]

i have no prior knowledge of AI. But, I'm good in data structures and algos, and comfortable in JS as well. Can i start from this playlist? Or do i need to study some pre-requisites?

[@Radu]

You probably can. Coding only starts later, anyway, after the AI things about Neural Networks.

[@MRX-nm5dn]

nyimak bro. Terimakasih. Thank You

[@Radu]

You're welcome :-)

[@MRX-nm5dn]

​​@@Radu let x = 0.19 let w = -1 let b = - 0.20 function carStop(){ if (x * w > b){console.log("stop that car") x = 0 } if (x * w < b){console.log("move that car")} } carStop() Homework 1 how many solution with same value (car stop in same position like on that video) the answer : we can multiply b with 2 and then multiply w with 2. and the car will stop in same position. homework 2 make the car bounce. answer = we can make x variable to 0 again in the first if block like this = let x = 0.19 let w = -1 let b = - 0.20 function carStop(){ if (x * w > b){console.log("stop that car") x = 0 } if (x * w < b){console.log("move that car")} } carStop()

[@MRX-nm5dn]

@@Radu and sorry i use console.log to visualize my answer like your video, coz my phone cant operate your app in your website.

[@Radu]

Oh, ok. Yeah, I didn't make the playground with a phone use-case in mind. Need a big screen and a mouse.

[@eric7758]

Hi Radu, thanks for your teaching style I appreciate everything. Could you do some teaching on autocoder and decoder maybe something simple like identifying shapes and lines from images. Using only JavaScript? I need to learn encoders work. Thanks

[@Radu]

Thanks for the idea, maybe someday... have to see how to fit those in my upload schedule.

[@KisuuleGodwin-uc4lx]

i truly give credit to this guy for his course is it possible to disturb forex trading and apply the same concept because you're the wright boss for all computer work

[@Radu]

I don't think anything happens to forex when I publish these videos :-))

[@catakuri6678]

25:06 My instinct is infinite solutions

[@Radu]

Good instinct :-)

[@charlesukanga4835]

Many of my friends never knew that AI and machine learning was possible with JavaScript, until I introduced them to Radu

[@Radu]

True, AI and ML are not something language dependent.

[@wilfredomartel7781]

😊 maybe a video neural networks with genetic algorithms?

[@Radu]

It's there already in Phase 1 :-)

[@fadyasaad1105]

I think writing the weights manually will be good, instead of creating a GUI for that, we could use the browser prompt dialog to get the new value by pressing double click on the component, (the neuron or the connection)

[@Radu]

It's a good idea. Double clicking and then typing something is probably less work than using the scroll wheel, but not significantly different and... need to use the keyboard. I sometimes teach by going through the desks with my mouse in my pocket and make updates on the screen from anywhere in the room :-D

[@fadyasaad1105]

@@Radu Yah I got you 😅. Maybe a popup slider dialog will be more helpful, with an input html element of type number.

[@Radu]

Maybe, yeah! Continue to check the next lessons. The playground editor becomes more powerful in lesson 2 already. Curious if you have more advice for me then.

[@fadyasaad1105]

@@Radu Can't wait for lesson 2 video! This content is absolutely amazing, keep up the great work! 😄

[@Radu]

@@fadyasaad1105 Cool :-) coming out on Friday.

[@srhtonr5917]

Hi Radu, I see that you have both Machine Learning, Self Driving Car and now this series. What courses would you recommend for a beginner to take ? Should I watch all of them?

[@Radu]

Do you know JavaScript already? If so, you can start with either of them, I think.

[@srhtonr5917]

Yeah i know javascript, i'm a javascript developer, but would you say starting from machine learning course and going through the self driving, and ai is the best way ? because im planning to apply fora masters degree on artificial intelligence and i want some fundamentals before going into that field, and i saw that you create your sessions with js, it just amazed me @@Radu

[@Radu]

@@srhtonr5917 I think the fundamentals you get from either of my courses will help you understand things much better in a masters degree on AI :-) you can start with any one of those, just prepare for a lot of coding since we don't use libraries.

[@srhtonr5917]

Thanks bro, keep it up good work, you are one of the most authentic content creator of the software/youtube@@Radu

[@rikki7663]

hey radu do i need to cover yur Self driving car-javascript course before starting this ML course

[@Radu]

Not really... The first 5 lessons have no JavaScript in them. It starts just after that. If you find it too difficult to follow along then, you probably have to start over.

[@omeralishin]

👌

[@Radu]

:-)

[@sagarkashyap3397]

hey radu, will you be posting one lesson every day, o one video a week..? i want to know so that i can plan my studying schedule.

[@Radu]

It will be one per week, every Friday, as usual.

[@disrael2101]

you didn't really touch how the sensors related to the x, weight and bias, also the neurons function now are simply if statement you didn't really touch how they work , learn etc.. anyway can't wait for the next video in the series thanks so much for sharing!

[@Radu]

The one sensor we played with (front sensor) is a variable (x). The weight and bias are 2 numbers. We can play with them in the playground, but they are considered fixed; meaning that a neural network, once trained, those values don't change anymore. Yes, the logic here can also be implemented using a simple if statement, and, actually the logic at the end of this course could be implemented as a bunch of if statements as well [ maybe a good practice :-) ]. Machine Learning just means that instead of setting those weights and biases by hand, we rely on some labeled data and some algorithm to set them for us so that the neural network works reasonably well. Here we're the algorithm.

[@disrael2101]

ok thanks that start to slowly make sense now, but still need lots more practice examples with ai and machine learning to actually grasp how to make the computer do all the manual work for us but i can clearly see how this is the most promising concept for our future as humans @@Radu

[@ChandrashekarCN]

💖💖💖💖

[@Radu]

:-)

[@armansaha297]

Radu Sir, can you explain why every game developer uses classes in the place of objects while the code needs to declare them only once. Like- constructor for main player or entire game.

[@Radu]

There can be many reasons... In some programming languages (like Java), you have to use Classes. Also... even if you can avoid it (like in JavaScript), you may want your code to be consistent (using similar structures all the time).

[@andradegilmar]

I am speechless

[@Radu]

Yeah... because KZbin doesn't support voiced comments :-D

[@Nothing-hr2ef]

Bro I need guidance for this video, I am beginner at programming, I know some basic JavaScript. Suggest a playlist to follow this tutorial.

[@Radu]

Hi, there's no JavaScript in this video or the next 3 lessons of this course. Later, there is... and I do have a playlist linked in the description that can help out, but if not enough, you can check JavaScript courses from FreeCodeCamp, for example.

[@priyanshubisht8326]

Sir how to start backend and from what resources

[@Radu]

I rarely teach back-end on this channel. I guess, first you have to ask yourself what language you want to use. JavaScript is good, if you know it already... and it can be used as the front-end as well. So, maybe looking for NodeJs tutorials is a good start?

[@animespiderevil4564]

bro i am new to javascript language can i follow this course or where i can study javascript please tell bro

[@Radu]

First 4 lessons in this course don't have any JavaScript so you can try to see how it is like. After that... if you think you need to study web development basics first, I have a beginner course on the HTML Canvas here on KZbin. It's the same technology we use in the self-driving car course. I also have some prerequisites in a playlist in this video description if that course is not enough. It goes through the math and coding fundamentals we use here. And finally... I recommend you start with Phase 1 of this course. I take things easier there and explain the code more in-depth. Hope this helps :-)

[@alwysrite]

will have to watch a few times to get the hang of it - once I get it then I will move forward - Get it ?- move forward ?...because the car was moving forward .?..wait ...no ... no ... no...

[@Radu]

:-)))) good one! *noting it down*

[@firefox_tn1815]

hi radu i think that you made this playlist public by mistake 😅

[@Radu]

Reading your comment here almost gave me a heart attack :-D I have no idea how you commented here, and how you found the video :-\ the playlist and the video still show as unlisted to me...

[@Radu]

Ah, you might have found it on Github... I thought I made that one private :-\

[@firefox_tn1815]

Yeah it's from GitHub. Don't worry I didn't watch it I don't want to spoil it. Good luck Radu

[@firefox_tn1815]

​@@Raduby the way there's another video that's still public

[@Radu]

@@firefox_tn1815 cool! I don't bother to hide it anymore since it comes out tomorrow :-)

[@sureshsingh9880]

Sir some about generative AI

[@Radu]

Maybe someday...

[@bonagirisaisantosh5408]

Radu, Can you please provide the schedule of this phase 3???

[@Radu]

My upload schedule stays the same, one video per week, every Friday at 16:00 (finnish time). Tomorrow lesson 2 comes out, then one lesson per week until mid-May, pretty much, unless something major happens.

[@bonagirisaisantosh5408]

@@Radu you have showed the playground in this video where we ended with different thing in phase 2 ... will any video in this phase gonna continue from that phase 2 ending one ?? Just out of curiosity😅

[@Radu]

Hi, I'm not sure I understand what you men.

[@bonagirisaisantosh5408]

@@Radu I mean the track that we have designed in the phase 2, isnt being used here...

[@Radu]

Ah, ok. I see :-) I did make the scenario we used so far using the world editor from phase 2 :-) few other scenarios will follow (next lesson). Then, in lesson 5, we will eventually use the world editor to create a new scenario and we will also use the one we created in phase 2 eventually, I think (I forgot, to be honest :-)))

[@wilfredomartel7781]

😊😊❤

[@Radu]

:-)

[@javifontalva7752]

🤯🤯🤯🤯🤯

[@Radu]

:-))

[@CA-at-YT]

Still have to finish part 2, but it seems to me that a next coffee delivery is already in the pipeline...

[@Radu]

Good luck :-) and thanks!

[@DanielJoseAutodesk]

I have a hard time understanding something in JS. Since I have more experience in C / C ++, I even got some JAVA books to try to understand something and I'm not succeeding. Here's the thing: In some of your code, I see the expression min = [...points [0]]; What this expression means ?!?! I imagine it must be to create an array. Or it would have another purpose, like copying the data from the array to another array ?!?!🤔🤔

[@Radu]

Hi, I'm not sure where you saw something like that. The closest thing I could think of is 'const min = Math.min(...points);' where you just spread the array of points to individual numbers (assuming points are 1D here) so the Math.min can take them as parameters. Or... 'const newPoints = [...points];' where we create a new array with the same content as the points array, but not a reference to the points array.

[@DanielJoseAutodesk]

@@Radu Thanks for the reply. Now I can understand what is being done. 😁👍 I saw this expression in the utils.js file line 74 of the Zip ml-course-main / P6.k_nearest_neighbors_lesson

[@Radu]

Ah, ok. I see.

[@yashpreet4103]

⚡⚡⚡👍

[@Radu]

:-)

[@abdularham4452]

will i be able to get a job after this course ?

[@Radu]

I didn't make this course with specific jobs in mind, but... You will learn many things about software engineering during this course, so your web development skills will improve and you will learn to manage growing projects. You will also learn basics of artificial intelligence, but you cannot really apply for AI jobs using just this. They typically use libraries that you need to master to get the position, and I don't use any here... Instead, I show how to implement several things from scratch so it is clear why things work, not just to show that they do.

[@jitulteron6542]

The God

[@Radu]

:-))

[@javifontalva7752]

Please a thumbs up if you find this vid amazing!!!

[@Radu]

I gave it a thumbs up, but I don't think it's 'amazing' :-)

[@londelidess]

KZbin gemmmmmmmmmm

[@Radu]

:-)

[@maxkosh4839]

I thought you are from Romania

[@Radu]

Originally I am :-)

[@user-qe4nu5sz7n]

hmm, I commented and comment disappeared

[@Radu]

It's possible... KZbin does weird things sometimes. I never removed any comment.

[@pizdaxyu]

bery nice!

[@Radu]

Thanks! :-)

[@user-px5pj7ux5k]

very nice sir,please create also a tutorial about the elastic collision of a line and a circle.😊❤

[@Radu]

Maybe someday :-) now I have content scheduled for the rest of the year.

[@MRX-nm5dn]

let x = 0.19 let w = -1 let b = - 0.20 function carStop(){ if (x * w > b){console.log("stop that car")} if (x * w < b){console.log("move that car")} } carStop() Homework 1 how many solution with same value (car stop in same position like on that video) the answer : we can multiply b with 2 and then multiply w with 2. and the car will stop in same position. homework 2 make the car bounce. answer = we can make x variable to 0 again in the first if block like this = let x = 0.19 let w = -1 let b = - 0.20 function carStop(){ if (x * w > b){console.log("stop that car") x = 0 } if (x * w < b){console.log("move that car")} } carStop()

[@Radu]

You seem to be re-implementing what the network is doing, in code :-) why? Also, note few things: 1. x is a variable (it should be a parameter to the function) 2. your naming "carStop" is opposite (in the video we talk about accelerating or not, not about stopping) 3. you do not handle the case x * w = b, but the network does 4. about answer to homework 1, so... you found one more solution, are there others? PS: Discord might be a better place to discuss these things, you can also send screenshots from the playground if needed to show something from there.

[@MRX-nm5dn]

@@Radu i re implementing what the network is doing coz my brain didnot enough to imagine the video instantly, and my phone didnot able to play your app in your web. answer for your note : 1. how to make x become parameter in the function? (i will find out too in the internet) 2. okay, i use carStop name coz i think in difference way but have same purpose like in the video (is this statement true?) 3. its true, i dont yet handle it. i will repair it (i still try to solve it) 4. another solution is we can devide b with 2 and then devide w with 2. and the car will stop in same position. and i just install discord to listen to you. thanks a lot.

[@Radu]

I see you edited and added an answer for homework 2 as well. Note that you cannot change the logic for that. The neurons always fire in the same way (when stimulated enough). What you need to change is w, b and w2, b2 (because now we have a second neuron). And remember that x is the variable (it changes as the car moves), the others stay the same once we set them.

[@Radu]

It is a good practice. Let's see how far you can go in the next lessons :-) You should be able to go like this all the way, recreating the logic with many if-else statements. Wonder if anybody will eventually try that! 1. Just pass it to the function 'function carStop(x) { ...' and call it like 'carStop(0.19);' 2. I understand, but I introduced that reverse control at the end, that is... intentionally stopping the car. Homework 1 is just letting physics (friction) stopping the car if no forward acceleration is applied. So, might make more sense to keep the word 'stop' for then, otherwise your code will become confusing to read. 3. Typically IF conditional statements are followed by an ELSE. I recommend using it here. The neuron either fires or does not, no need for 2 IF statements. 4. Ok, you found another! But why multiply and divide by '2'? Couldn't you use other values there?

[@rad1m]

For the song!

[@Radu]

Wow thanks :-)) maybe I should start a singing channel!

[@tanaymishra4483]

Sir i don't even wear hats but still hats of to you.. I can't express how great you are in words.. It's impossible to find someone these days whose central idea of existence in not Money.. In slags one might/can call you gigachad...🫡😅

[@Radu]

Thanks :-D