- Видео 9
- Просмотров 140 259
Challenging Luck
США
Добавлен 28 июн 2020
I sometimes solve problems
Product Classification - NLP - Save 10,000 Labor Hours
In this video I classified products based on their description for e-commerce sites.
Jupyter Notebook and Dataset: github.com/challengingLuck/youtube/tree/master/ML%20For%20Entrepreneurs/Product%20Classification
For project or tutoring requests contact me at: challenging.luck.business@gmail.com
I have an Instagram now: challengingluck
#machinelearning #entrepeneurship #naturallanguageprocessing #nlp
Music: www.bensound.com
Jupyter Notebook and Dataset: github.com/challengingLuck/youtube/tree/master/ML%20For%20Entrepreneurs/Product%20Classification
For project or tutoring requests contact me at: challenging.luck.business@gmail.com
I have an Instagram now: challengingluck
#machinelearning #entrepeneurship #naturallanguageprocessing #nlp
Music: www.bensound.com
Просмотров: 1 191
Видео
Demolishing Battleship Board Game Using Probability - Building a AI in Python
Просмотров 4,4 тыс.3 года назад
The code is written in Python. I included all three bots in the code in GitHub. The final bot is the cleanest and most well-documented one since that was the one that could actually perform well. Code on GitHub: github.com/challengingLuck/youtube/tree/master/Battleship Click here or I will sick Mark Rober on you: bit.ly/2Z1uzKM If you want to send me a tip. You absolutely don't have to! Etheriu...
Dominating Monopoly Using Markov Chains
Просмотров 12 тыс.3 года назад
Github: github.com/challengingLuck/youtube/tree/master/monopoly Click here or I will sick Markov on you: bit.ly/2Z1uzKM If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb Music: www.bensound.com
Backtracking Explained - Solving N-Queens and Knight's Tour using Python
Просмотров 34 тыс.3 года назад
Github: github.com/challengingLuck/youtube/tree/master/backtracking Click here or I will steal your cookies: bit.ly/2Z1uzKM If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb Music: www.bensound.com
Simulated Annealing Explained By Solving Sudoku - Artificial Intelligence
Просмотров 85 тыс.3 года назад
Code - github.com/challengingLuck/youtube/blob/master/sudoku/sudoku.py Check this out for good luck: bit.ly/2Z1uzKM Paper on which my work was based on: link.springer.com/content/pdf/10.1007/s10732-007-9012-8.pdf If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb Music by Vincent Rubinetti Download the music on Bandcamp: vincerubinet...
The Birthday Paradox - Simulating 10,000 Times
Просмотров 3864 года назад
Code: github.com/challengingLuck/youtube/blob/master/probability_problems/birthday_paradox.py Don't click this: bit.ly/3tAtffM If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb Art from: www.freepik.com Designed by pikisuperstar
Re-rolling Dice At 1 VS 6 - Simulating 100,000 Times
Просмотров 7564 года назад
Code: github.com/challengingLuck/youtube/blob/master/probability_problems/roll_at_6_vs_1.py If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb Art: www.freepik.com Designed by vectorpouch
The Monty Hall Problem - Simulating 100,000 Times
Просмотров 1,2 тыс.4 года назад
Code on Github: github.com/challengingLuck/youtube/blob/master/probability_problems/the_monty_hall_problem.py If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb
The Lost Boarding Pass - Simulating 100,000 Times
Просмотров 1,4 тыс.4 года назад
Code on Github: github.com/challengingLuck/youtube/blob/master/probability_problems/lost_boarding_pass.py If you want to send me a tip. You absolutely don't have to! Etherium: 0xae6b07d67301d4f5aed36254b19f05cc593732cb
good day, what are the other backtracking algorithm that i may use to solve the local maxima of the steepest ascent hill climb? instead of using the depth-first search, can you suggest another backtracking?
Bruh Of course its 66/33 if you lose when the car door being opened by the host is included. It has nothing to do with the problem tho. If the problem is only opening the door on a goat and then having to switch or stay then its 50/50 no debate. Whats the point of simulating the games where the car door gets opened and u lose no matter if you switch or stay? Its not a problem its just retarded
Simulated congealing at the low temperature phase. :) Great video, thanks!
Why do you choose this music for this video? Are you teaching how to use a sword?
lol
It was not clear which point was x1 & x2 @ 2:20, when you look at youtubes replay-graph it seems I was not alone :)
"Tap the like button gently" 😀
thank you, you helped me a lot for my assignment !
merciiii infiniment
Awesome video!
I do understand the thinking of the algorithm and the code is amazing. But I am quite confused that why the board metric could change during the function 'solve'. As it is a parameter for the function which the return is a bool variable True or False. Why would the argument 'board' be changed during the running of function? I am curious.
This looks like the same animation style as 3 blue 1 brown. Is there a specific software you use?
Python library in python called Manim
Great video! thank you so much!
good one dude
can somebody explain why multiplying all these matrix give the "final transition matrix" ? And I don't understand the code line : steadyStateVec= steadyStateVec/sum(steadyStateVec) . Is there a theorem behind it ?
Hi, where can I contact you ? I'd really like to ask you some questions
do you remember how long your knight's tour algorithm took
A few seconds I believe. The code is linked here. You can just try running it.
love this
I love how you talked about finding the initial temperature, this is often overlooked.
Thanks for noticing :)
does this find a path for every starting point
yep
what is the best way to learn ML/ data stuff? any resources. Спасибо
I’d start with RUclips and what’s free
Assume a column or a row is like 1 2 3 2 5 6 2 8 9. What's the cost of the column or the row?
The cost would be 3.
@@ChallengingLuck shouldn't the cost be 2 considering there are 2 duplicates of "2" and 1 is original.
How simple and great explanation it is... Very interactive!!
Thank you! 😃
Would have been nice to see a comparison between the steps in simulated annealing vs. brute force backtracking. Brute force backtracking can also solve sudoku very quickly since it isn't that hard of a problem space
Backtracking is computationally faster than SA for solving sudo.
Good explanation. Visualisations are fantastic
Thank you dude. Your video is short and to the point. 🎉❤
Is there a evidence that this is better than just local search algorithm
Didn't make any such claims.
This video was too hard to follow, dropped off at 2:10 👎
Reach out to me and I’ll give you a free 1h session where I explain SA.
Great videos
Thanks!
Another great video!
Glad you enjoyed it!
👏very clear!
Glad it was helpful!
Me: mom can we have CodeBullet Mom: no we have CodeBullet at home CodeBullet at home:
Ouch!
@@ChallengingLuck hey at least you upload videos
Is there a reason why you don't also sum the duplicates in each 3x3 square of the sudoku? A valid solution with distinct numbers in each column and row does not guarantee a sudoku solution, correct? Also at 0:57 shouldn't the temperature graph have # of states checked on the x-axis? This confused me a little bit, but Wikipedia set me straight. Regardless, it was a pretty helpful video, and I liked the 3b1b music :)
When the 3x3 squares are randomly built, I make sure that there are no duplicates. That's the reason why I only check rows and columns for duplicates.
Ah ok, makes sense.
Muy bien explicado Mister Luck.....saludos de los andes peruanos
Thanks for the comment. Glad you found it helpful.
Thankyou for your sharing. It is helpful to me. :)
Great to hear!
Thank you
You're welcome
Class video , thanks
No problem 👍Thanks for the kind words.
You never explained at what point the temperature is lowered and by what amount. Also, how do you get the initial temperature value?
Thanks for the question! The temperature is continuously dropping from the moment the algorithm starts. As the algorithm runs, the temperature drops at a constant rate. I explain how I set the starting temperature at minute 4:33.
I will post my algorithm based in backtracking + restrictions to solve sudokus in python: github.com/dogia/sudoku_solver
Very clean! Good job
greetings, in the knight tour example, how did we choose the 3,2 as the starting point? i mean shouldn't we iterate through the starting points as well? thanks in advance
Depends on the approach!
This code is niccccce. Big ups to you guys. Freakin brilliant
Thank you so much.
Love it. :D
Thanks for the comment.
Brilliant explanation!
Thank you!
very good explanation
Thanks for liking
Your model look suspiciously like Code Bullet lol anyways great video
True. I like his videos
Can you also make a video on ship placement. Since beating your static bot is pretty easy, junst dont place any ships in the center. So you need also some randomness in firering mode. Do you think, you can come close to the nash equilibrium?
Could be a future video. Let me know if you are interested in working together.
@@ChallengingLuck I would like to see your future video. I assume you can tweak your generator to give it a tendency to place ships more often (not always) at or close to the border. I guess, that if you get a generator and let it run 1.000.000 times, the propability of each tile beeeing a ship should be roughly similar. Of course, I might be wrong since I did not calculate the nash equilibrium.
I think a mayor safe improvement is to calculate the tiles where there can not be a part of a ship. For example, if you already fired 4 shots around a tile and all of them missed, there can not be a ship because the smallest ship is size 2. Of course, if the smallest ship is already sunk, you can exclude more tiles. This method should be a simple to implement improvement and all other techniques, no matter if its a neural network, Monte-Carlo, Random or the diagonal method benefit from it.
I really like your comment. Let me know if you want to collaborate.
@@ChallengingLuck Thanks, for a collaboration I do not have the time. However, I suggest we can discuss here, so its open for everyone. Again, my main point is, that first the shooting player must find out all useless fields to shoot, this is the easy part and can be solved pretty quickly, if you look for all possible combinations for the smalles undetected ship. The hard part is, to come up with a better strategy than the diagonal one.
Very cool. Short, well written script
Thanks for the kind words.
Why don't you simply reject the actions/flips that increase the system energy and accept the ones that decrease the energy? Why use the probability (i.e. shouldn't the probability just be 100% and 0%)?
Good question. The reason for that is because even though some actions might reduce the energy you could reach a local minimum, not a global minimum. In the Sudoku case you could reach a point where you only have 2 errors let's say but in order to solve the puzzle you need to shuffle more than just those two numbers. In that just going with the option that decreases energy would get you stuck. Hope this helps.
@@ChallengingLuck Ah makes sense. Thank you.
how would you implement this for sudoku
That's for you to find out, isn't it
Wow I was impressed. Simple video, to the point. Maybe do something unique to make it stand out more, but I like your presentation style
I appreciate that!