Lognormal is extremely useful when analyzing stock prices. As long as the growth factor used is assumed to be normally distributed (as we assume with the rate of return), then the lognormal distribution makes sense. The normal distribution cannot be used to model stock prices because it has a negative side, and stock prices cannot fall below zero.
abhijeet if you have two prices of an asset delta_t apart S(t) and S(t+delta_t) in time then ratio S(t+delta_t)/S(t) can be modeled as a e^(r*delta_t) if asset grows with 100% certainty at rate r [here e is exponential e~2.718281828] ... when 100% certainty is not there then there is a variance term in exponent of e ... in that case ratio could be modeled as e^(r*delta_t+v*delta_t*randomness) ... see this page on wikipedia: en.wikipedia.org/wiki/Geometric_Brownian_motion
Hello Dhaval Sir,please keep uploading videos of this series , as for beginners of data science students, your simple explanation is very helpful, thanks
Appreciate the way you answer the “so-what?” questions, Dhaval. That’s one aspect of ML/DL instruction most tutorials frequently miss out out on. Kudos!
Amazing, thank you very much! I was studying the Log function in Math because it's used in Data Analaytics. I just couldn't clearly understand where I could use this. You video helped make it all clear to me now! Thanks a lot =D
Hi Dhaval I hope you are doing good and safe. This is what I love in your teaching Dhaval, your explanation is super smooth, anyone can understand this. Generally who are afraid of maths and stats I recommend them to watch codebasics maths and stats playlist. Thanks for this initiative Dhaval.
In case of the credit approval model example, why don't we remove Poojas data since her income is definitly an outlier? what are the scenarios we use log value vs remove outliers?
When we use data is normal distributed, does it mean that all values are almost same range or same scale. 8n other words, not have outliers or data is not wide spred out or some data is almost similar than one big data and cannot compare them. If data does not have any problem like that then we say the data is normally distributed. And data is already cleaned and filled. Am I right?
So if my data do not meet the linear regression model, I just change the data to make them fit better to the model, and not the model that might explain the observation.
Thank you for this playlist.. a quick question, will you upload more videos on this playlist ? or can you suggest a place where I can learn other stats topics like Probabilities and hypothesis testing?
Sir why don't we just use standard scaler what is the difference between the two?! + I can't get why we need to convert the data to the normal distribution in the first place!
can you please explain what is the actual use of this distribution, I just know only by using Box-cox transformation we can convert it into gaussian distribution. is there any other property of this distribution like Gaussian distribution..?
How can I find mean and variance of probability distribution using python? I have mentioned a problem below. Thanks in advance! The number of times I go to the gym in weekdays, are given below along with its associated probability: x = 0, 1, 2, 3, 4, 5 f(x) = 0.09, 0.15, 0.40, 0.25, 0.10, 0.01 Calculate the mean no. of workouts in a week. Also evaluate the variance involved in it.
Log normal distribution or log scale is used for data visualization so that the outliers doesn't suppress rest of the data points. How do you treat those data points with higher values totally depends on you, you can remove them as outlier or use some other techniques to treat them
Lognormal is extremely useful when analyzing stock prices. As long as the growth factor used is assumed to be normally distributed (as we assume with the rate of return), then the lognormal distribution makes sense. The normal distribution cannot be used to model stock prices because it has a negative side, and stock prices cannot fall below zero.
abhijeet if you have two prices of an asset delta_t apart S(t) and S(t+delta_t) in time then ratio S(t+delta_t)/S(t) can be modeled as a e^(r*delta_t) if asset grows with 100% certainty at rate r [here e is exponential e~2.718281828] ... when 100% certainty is not there then there is a variance term in exponent of e ... in that case ratio could be modeled as e^(r*delta_t+v*delta_t*randomness) ... see this page on wikipedia: en.wikipedia.org/wiki/Geometric_Brownian_motion
both of you are correct. reminds me of my days at School of Mines.
Hello Dhaval Sir,please keep uploading videos of this series ,
as for beginners of data science students, your simple explanation is very helpful,
thanks
Appreciate the way you answer the “so-what?” questions, Dhaval. That’s one aspect of ML/DL instruction most tutorials frequently miss out out on. Kudos!
wow, the way he explicitly communicated the meaning of a log-normal distribution, so many lecturers keep confusing me until now. Thank you so much
Your classes are amazing, anybody can understand with examples you are taking
Thanks for your great explanation! Really helpful, brief, and to the point.
I just came across this problem and your video helped me to understand the concept! Thank you very much!!
Please sir complete this series as much as possible .
Thank you for all your efforts
Thank you for explaining the concept of log normal distribution well.
Amazing, thank you very much! I was studying the Log function in Math because it's used in Data Analaytics. I just couldn't clearly understand where I could use this. You video helped make it all clear to me now! Thanks a lot =D
Please complete this playlist as soon as possible. It will very helpful for the learners.
Linear Algebra and Calculus course next too sir?😁
I mean ofcourse after this series is fully completed sir...
Hi Dhaval I hope you are doing good and safe. This is what I love in your teaching Dhaval, your explanation is super smooth, anyone can understand this. Generally who are afraid of maths and stats I recommend them to watch codebasics maths and stats playlist. Thanks for this initiative Dhaval.
Thanks chinmaya for your kind words. Hope you are doing well buddy ☺️🙏
Really appreciate your great work and simple explanation! I now have a better understanding.
Glad you liked it ! 😊
Please do a video per day so you can easily educate many Noobs to data science experts ..
In case of the credit approval model example, why don't we remove Poojas data since her income is definitly an outlier? what are the scenarios we use log value vs remove outliers?
very well explained. your examples are awesome. very practical and easy to understand. keep up the good work
Glad it was helpful!
Completed this playlist..Really great. thank you so much.eagelry waiting for your next videos
first time in my life that i understand this
Simple and easy to understand. Thanks !
Sir, can you make a video on the important topic of ML,Statistics and Deep learning for the data Scientist ?
Fantastic video. Most intuitive explanation online. I subscribed.
Great video! Most of the data here was right skewed..so this log normal transformation also works for left skewed?
I don't think so .....it would result in an even more left skewed distribution
Can we use minmaxScaler or standardscaler instead of log?
Thanks fpr your time!! good explanation :)
awesome explanation!!!
When we use data is normal distributed, does it mean that all values are almost same range or same scale. 8n other words, not have outliers or data is not wide spred out or some data is almost similar than one big data and cannot compare them. If data does not have any problem like that then we say the data is normally distributed. And data is already cleaned and filled. Am I right?
good explanation...One query! Don't we need to transform age variable also because now age is dominating as compared to lognormal income variable?
Log normal is also useful in measuring Claim Size for insurance company as it has range from 0 to infinity
Very nice, thanks!
love the explanation, so easy to understand
Glad it was helpful!
So if my data do not meet the linear regression model, I just change the data to make them fit better to the model, and not the model that might explain the observation.
sir is this much mathematics sufficient for absolute fresher in the field of data science.
Thank you this was clear.
Hi do you implement any hybrid machine learning models? collaborative filtering, rule based classification, and association rule coding....
Thank you for the great video!
Dhaval sir make video on probability and conditional probability wich is also part of statistics and used in data science...
Hi,It is possible generating random variables in excel from truncated normal distribution?
sir when should i do logy and when logx
Thank you for this playlist.. a quick question, will you upload more videos on this playlist ? or can you suggest a place where I can learn other stats topics like Probabilities and hypothesis testing?
Sir why don't we just use standard scaler what is the difference between the two?! + I can't get why we need to convert the data to the normal distribution in the first place!
Nice explanation
Really good lectures. Can we have full playlist on statics?
Thank you sir, good demo
Sir, great content. Is log normal distribution part of normalization method in machine learning?
Please make more videos. All are good
Yes I will
can you please explain what is the actual use of this distribution, I just know only by using Box-cox transformation we can convert it into gaussian distribution. is there any other property of this distribution like Gaussian distribution..?
Hello Dhaval sir, this video helped me understand the concept in a better way. Also, could you please upload a video on Random variables.
Can you please make a video on ANOVA?
Hello sir, please activate ads for your video, i was jumped to other video in this playlist, but i dont find any ads :)
Pooja, what is this behaviour????? ❤
How can I find mean and variance of probability distribution using python? I have mentioned a problem below. Thanks in advance!
The number of times I go to the gym in weekdays, are given below along with its
associated probability:
x = 0, 1, 2, 3, 4, 5
f(x) = 0.09, 0.15, 0.40, 0.25, 0.10, 0.01
Calculate the mean no. of workouts in a week. Also evaluate the variance involved in
it.
thank you so much😍
very informative
Fantastic explanation here, thank you so much. Really clear and simple explanation of the concept.
Thanks for such a nice video.
I am glad you liked it
GREAT WORK!!!!
Thank you.
good class
Does anyone know when we should use log normal distribution instead of remove the outliner?
Log normal distribution or log scale is used for data visualization so that the outliers doesn't suppress rest of the data points. How do you treat those data points with higher values totally depends on you, you can remove them as outlier or use some other techniques to treat them
What are the moments?
Plz teach arima model
Why do we need bell curve?
because the average median has more sense, a lot of statistical concepts can be applied to that standard distribution
nice
If possible do a video per day
I want to take tutorial
Dislike for the late post. Humble request to you to upload videos more frequently.
Gagan, I have 9 to 6 job and I get time only on weekends for this. Making these videos take lot of time, hope you understand
@@codebasics Completely understand. Really appreciated the content. You are making learning easy for us.
-My wife works in a hospital.
**Thought now I am gonna get some serious data**
-She says some people spend MANY days.
Hwy Indians use "You Knowhh" in every line, it feels very irritating to hear
nice