I'm a commerce student. I started watching your videos because of time series analysis of stock prices. Your data science are still so easy for me to understand.
Deeply appreciate the importance of choosing loss function for different models. As a beginner in data science it makes a lot of sense and better understanding of concepts.
I really love the way you summarize a lot of the concepts, this is helping me a lot to understand other lectures in some of my courses. Great Job! Many thanks!
Hi! Is there a way to choose a loss function from a predefined list or even to define a loss function by myself to use it with scikit-learn algorithms? Or do I have to write a whole ML algorithm from scratch to change default loss function?
Thank you. For the square loss, why don't we heard about cubic loss or quartic loss? Also is the square loss choice also based on its derivative compared to a linear loss?
Squares are used to avoid negative values. Also the square increases the range, penalizing the worse values. Cubic (or any odd power) will turn negatives. Using hight positive powers might hurt discovery and/or make analysis hard (the values will be too big or small depending on domain). It also gets expensive
Is it possible to apply a custom loss function in a regression model (or any other algorithm for predicting continuous variable) ? I'm working on a stock market prediction model and I need to maximize the following loss function: if [predicted] < [actual] then [predicted] else [-actual]. Would that be possible ? Thanks
@@pizzaguy8484 it isn't indeed haha, but i saw no comments and then commented, i'm a latin one, and i prefer english content because im learning tghe language, you comment really helped me , thank you so much
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
This is such an undervalued channel. You are out here making data scientists out of everyday people. Thank you for your content.
I'm a commerce student. I started watching your videos because of time series analysis of stock prices.
Your data science are still so easy for me to understand.
This brought it all together so nicely. You're a great teacher
Thank you so much!
Deeply appreciate the importance of choosing loss function for different models. As a beginner in data science it makes a lot of sense and better understanding of concepts.
This is the best explanation of Loss function among all blogs, videos I came across
I really love the way you summarize a lot of the concepts, this is helping me a lot to understand other lectures in some of my courses. Great Job! Many thanks!
Great to hear!
Exceptionally gifted with intelligence, and the ability to understand subjects and transfer knowledge. Thank you!
waw! You are a great data science who conveys the ideas!!!!!!
I learn a lot with your videos. Simple and straigh to the point. Thanks for contributing my dear!
Wath an amazing tutorial you managed to give a detailed yet simple explanation, gained a subscriber
Ritvikmath ... you are a great communicator... thanks
I really like your videos! Just followed your channel last night and I am looking forward to more of your future videos!
Going through Data Science bootcamp at Flatiron School. This was supa' helpful. Thank you!
Complete new to ML and this was just amazing. Thank you!
Great to hear!
Very clear presentation. Thanks for keeping it simple.
yoo this was exactly what I needed! You speak so clearly too thanks man
Glad I could help!
You're amazing! Keep up the great work!
Another great video. Well done!
Thank you! Cheers!
absolutely amazing, your understanding and your explanation, both
Just a suggestion. a series on stochastic processes and possibly even short rate models.
really well explained great teacher!
Glad it was helpful!
Amazing video ! Nicely explained .
Amazing explanation
Please explain what is an ACTIVE Function, you are great!!!!
Amazing Video
very helpful for my exam. thanks
PERFECTLY explained! Thanks🙌
Glad it was helpful!
How would the minimum of weights to a hinge loss function be determined?
do i use 0-1 Loss if I am using a DecisionTreeClassifier?
Amasing video, thank you!!!!
great video！！！awesome！
thank you ... great explanation
Glad it was helpful!
Great explanation! Thank you so much!
Hi! Is there a way to choose a loss function from a predefined list or even to define a loss function by myself to use it with scikit-learn algorithms? Or do I have to write a whole ML algorithm from scratch to change default loss function?
nice examples!
Thank you, really good explained. You helped me to understand my lecture
Thank you. For the square loss, why don't we heard about cubic loss or quartic loss? Also is the square loss choice also based on its derivative compared to a linear loss?
Squares are used to avoid negative values. Also the square increases the range, penalizing the worse values. Cubic (or any odd power) will turn negatives. Using hight positive powers might hurt discovery and/or make analysis hard (the values will be too big or small depending on domain). It also gets expensive
incredible !!!! so good!!!!!
Perfect! Thank you
Can the loss functions be applied to all model indiscriminately? For example, exponential loss function to Support vector classifier? Thanks.
very nice lecture
Is it possible to apply a custom loss function in a regression model (or any other algorithm for predicting continuous variable) ? I'm working on a stock market prediction model and I need to maximize the following loss function: if [predicted] < [actual] then [predicted] else [-actual]. Would that be possible ? Thanks
Thanks, very helpful. Could you please do a video on Structural VARs (SVAR)? I'm really struggling with this concept. Greetings from Holland!
thank you
we need The practical example with python
Amazing
Hi.. can you make one video on how a white noise effect model while NLP algorithms are used
Thanks
That catch in the beginning😂😂
thank u teacher
POV: youtuber puts in ten times the effort to explain a concept compared to your university professor
OMG this is so useful
awesome vid!
In this video f(xi) is called the score
and exp (-yi f(xi)) is called the loss,
so what is y1 (f(xi) called? Does it have a name?
thank you
Thanks
almost let the pen fall on this one hahahah
haha!
I'm starting in data sciense, i'm not bad at math, butg my question has always been "how do i convert non-number data into number-data?, im noob 😴
you should check one-hot encoding, but I don't think it's related to this video at all.
@@pizzaguy8484 it isn't indeed haha, but i saw no comments and then commented, i'm a latin one, and i prefer english content because im learning tghe language, you comment really helped me , thank you so much
@@javierargueta7355 No worries! Good luck with your studies, don't give up!
Data science is a very vast topic . Any suggestions how to master it