search

kundtx / lfd2022-comments

0 stars 0 forks source link

Learning from Data (Fall 2022) #28

Open kundtx opened 1 year ago

kundtx commented 1 year ago

http://8.129.175.102/lfd2022fall-poster-session/16.html

ljh430 commented 1 year ago

G38 Jiahao Lai:It seems that alpha may change the distribution of the noise,but does different alpha change variance of the noise(which may indicate severness of the noise)?

peace-andjoy commented 1 year ago

@ljh430 G38 Jiahao Lai:It seems that alpha may change the distribution of the noise,but does different alpha change variance of the noise(which may indicate severness of the noise)?

G16 Xueqiong Yuan: Thank you for your question. Actually when alpha<2, alpha distribution do not have finite variance, but there is a scale parameter controlling the dispersion of the distribution. I think maybe you are asking if different alpha change its dispersion? We have set the scale parameter as a fix value, so the dispersion keeps the same with different alpha.

uprightman47 commented 1 year ago

G3 Siqi Chen: I think your work is very useful and interesting! I am wondering that where the alpha-stable noise will happen in real life. Could you give me an example? And how your contributions will benefit the practical applications? :D

peace-andjoy commented 1 year ago

@uprightman47 G3 Siqi Chen: I think your work is very useful and interesting! I am wondering that where the alpha-stable noise will happen in real life. Could you give me an example? And how your contributions will benefit the practical applications? G16 Xueqiong Yuan: Thank you for your question! For example, atmospehric noise is alpha-stable. I think our research does not have direct benefit to application, but it can provide some inspiration for theoretical research related to property of neural network and its generalization.

uprightman47 commented 1 year ago

@peace-andjoy

@uprightman47 G3 Siqi Chen: I think your work is very useful and interesting! I am wondering that where the alpha-stable noise will happen in real life. Could you give me an example? And how your contributions will benefit the practical applications? G16 Xueqiong Yuan: Thank you for your question! For example, atmospehric noise is alpha-stable. I think our research does not have direct benefit to application, but it can provide some inspiration for theoretical research related to property of neural network and its generalization.

Thank you, I've got it! Your explanation helps me to understand the neural network deeper and your work's theorectical value. ( •̀ ω •́ )y