Live Day 5- Advance Statistics With Python In Data Science

Поделиться
HTML-код
  • Опубликовано: 17 ноя 2024

Комментарии • 107

  • @upeshjeengar
    @upeshjeengar 10 месяцев назад +31

    🎯 Key Takeaways for quick navigation:
    06:34 📉 *Type 1 Error: Rejecting the null hypothesis when it is true leads to a Type 1 error, often depicted as a false positive.*
    13:13 📈 *Type 2 Error: Accepting the null hypothesis when it is false results in a Type 2 error, commonly seen as a false negative.*
    25:32 📈 *Two-Tailed Test: When investigating if a college's placement rate is different, it's a two-tailed test, checking both greater and lesser than the standard rate.*
    27:35 📉 *One-Tailed Test: If the question changes to checking if the placement rate is greater than 85%, it becomes a one-tailed test, focusing only on the greater side.*
    32:00 🎯 *Point Estimate: In inferential statistics, a point estimate is the value of any statistic that estimates the value of a parameter, providing an approximation of the population value based on sample data.*
    32:27 📊 *Point estimate is the value of any statistic that estimates the value of a parameter, like estimating the population mean.*
    33:30 📏 *Confidence interval is determined by the formula: Point estimate ± Margin of error. Margin of error accounts for uncertainty in estimating the population mean.*
    35:13 📈 *Example problem: Given population standard deviation, sample size, and mean, construct a 95% confidence interval using the z-test formula.*
    36:33 📉 *When population standard deviation is known, and sample size is greater than or equal to 30, use the z-test formula for confidence intervals.*
    49:48 🧮 *When population standard deviation is not given, and sample size is less than 30, use the t-test formula for confidence intervals.*
    51:05 📚 *Example problem: Given sample standard deviation, sample size, and mean, construct a 95% confidence interval using the t-test formula.*
    59:11 🎉 *Completion of confidence interval explanation and problem-solving using z-test and t-test.*
    [01:07:58 URL](ruclips.net/video/PbguWyuZ4Ts/видео.html) *📊 When conducting hypothesis testing, the third step involves stating the alpha value, often set at 0.05.*
    [01:08:28 URL](ruclips.net/video/PbguWyuZ4Ts/видео.html) *🤔 In the fourth step, defining the decision rule is crucial. For a two-tailed test with an alpha of 0.05, the critical values are ±1.96 for a Z-test.*
    [01:10:50 URL](ruclips.net/video/PbguWyuZ4Ts/видео.html) *🧮 Calculating the Z-test statistics involves the formula (X - μ) / (σ / √n), where σ is the standard deviation. Always use root n for sample data to account for the standard error.*
    [01:16:39 URL](ruclips.net/video/PbguWyuZ4Ts/видео.html) *🚫 If the calculated Z value falls outside the range of -1.96 to +1.96 based on the decision rule, reject the null hypothesis.*
    [01:28:29 URL](ruclips.net/video/PbguWyuZ4Ts/видео.html) *🧠 Applying the learned concepts to a practical problem, rejection of the null hypothesis indicates an increase in intelligence in this scenario.*

  • @AlexFamilyBonding
    @AlexFamilyBonding 11 месяцев назад +3

    Bank - ATM Opening in Given location -- 1. Collect the Population Mean - Personas Visiting the ATM - 2. Select the Sample Mean - typical > 30 sample persons vesting particular ATM 3. From the Population , we can find the std deviation 4. Based on the alpha - we can find the confidence intervals 5. calculate Z- Value 6. Based on the z -value - we can accept/reject the Null hypothesis

    • @GauravSingh-he1ne
      @GauravSingh-he1ne 10 месяцев назад

      I completely agree to you and I was thinking the same way as you but I got stuck at the situation that what would be the null and alternate hypothesis in this case? Can you help me by answering that?

  • @shelby1999
    @shelby1999 2 года назад +39

    Data science feels so exciting when you teach Krish!

  • @rajapratappedapudi
    @rajapratappedapudi Год назад +2

    After watching all statistics 7-day live sessions and now I have got so much confidence in stats, and I am very sure that I could learn data science subject. The way you teach is very clear and absolutely very easy to understand which I didn't find in my university. A Big Thanks to you Krish Naik ji for your great effort to help me and many students with your knowledge to shine their future and it is much appreciated. 🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏

    • @aryanthombre2971
      @aryanthombre2971 9 месяцев назад

      how are you preparing for data science

  • @kuldipdhokiya7733
    @kuldipdhokiya7733 2 года назад +6

    first i was doing petroleum engineering and one day I cam on this channel now I am preparing for data analyst/scientist job. :)

  • @ashwinmanickam
    @ashwinmanickam 2 года назад +18

    19:00 1 tail and 2 tail test
    49:00 t test- when to use
    54:49 degree of freedom
    1:04:08 Problem

  • @italiandarthvader
    @italiandarthvader Месяц назад

    Best video explaining p values and clear explanation of one and two tailed tests. Great work.

  • @harikishan80
    @harikishan80 Месяц назад

    Thanks for the video lectures. You have very clearly explained the concepts from basic to advanced level. 👍. Thanks for all your efforts. Much appreciated

  • @devilgod2980
    @devilgod2980 6 месяцев назад

    06:34 📊 Type 1 error occurs when we reject the null hypothesis when it's actually true, leading to a false positive conclusion.
    14:28 📉 Type 2 error happens when we fail to reject the null hypothesis when it's actually false, resulting in a false negative conclusion.
    19:04 📝 In a hypothesis test, a one-tailed test focuses on one direction (e.g., greater than or less than), while a two-tailed test examines both directions.
    26:15 🧠 Understanding the context of a question is crucial to determine whether a hypothesis test is one-tailed or two-tailed.
    32:41 📏 Point estimate in inferential statistics involves using sample data to estimate population parameters, recognizing that the estimate may not be exact but approximate.
    33:01 📊 Point estimates in statistics provide values that estimate parameters, such as the mean, from a sample.
    33:30 🛠 Confidence intervals are used to estimate the population mean, typically calculated as point estimate plus or minus the margin of error.
    35:56 📈 When given the population standard deviation, the z-test formula is applied to calculate the confidence interval.
    37:01 🧮 Given population standard deviation, sample size, and confidence level, constructing a confidence interval involves calculating the margin of error using the z-test.
    51:23 🧪 When population standard deviation is unknown, the t-test is used to calculate confidence intervals, with the sample standard deviation and sample size.
    01:09:26 📊 Z-test helps determine if a value falls within a specific range by comparing it with critical values like 1.96 for 95% confidence.
    01:11:05 📈 Z-test formula: (x - μ) / (σ / √n). Standard error (σ / √n) becomes relevant for larger sample sizes to match population means.
    01:15:06 🧮 Test statistic computation for Z-test: x̄ - μ / (σ / √n), considering sample mean, population mean, and sample size.
    01:16:53 🧠 Rejecting the null hypothesis in hypothesis testing suggests an effect, like increased intelligence due to medication.
    01:26:41 📉 T-test formula for test statistic: (x̄ - μ) / (s / √n), with x̄ as sample mean, μ as population mean, s as sample standard deviation, and n as sample size.
    01:28:29 🎓 Rejecting the null hypothesis indicates an increase in intelligence, highlighting the importance of hypothesis testing in real-world scenarios.

  • @harshmeshram4308
    @harshmeshram4308 Год назад +2

    🎯 Key Takeaways for quick navigation:
    20:25 📊 Two-Tailed Test: Checks if the rate is different, greater, or less.
    29:19 🎯 Confidence Interval: Uses point estimate, a statistic, to estimate a parameter within a range.
    30:19 📊 Confidence Interval Basics: Calculations involve finding upper and lower bounds to estimate parameters.
    32:27 📊 Point estimate is the value of any statistic estimating a parameter, like the mean.
    33:30 📏 Confidence interval: formula - point estimate ± margin of error, providing a range around the mean.
    35:13 🎯 Given a problem statement, population standard deviation, sample size, alpha, and mean, construct a confidence interval using the Z-test formula.
    50:19 🔄 If population standard deviation is not given, use a T-test for the confidence interval formula, even if the sample size is less than 30.
    01:08:28 📊 Specify decision rule in hypothesis testing considering alpha and tail type.
    01:09:06 🧪 Two-tailed test with alpha 0.05: critical values are ±1.96 for a standard normal distribution.
    01:11:19 📈 Standard error in z-test formula: divide by the square root of the sample size (root n).
    01:14:11 📝 Z-test formula for test statistics: (x - μ) / (standard deviation / root n).
    01:16:26 🤔 Rejecting the null hypothesis in a z-test implies the medication had an impact; direction depends on the z-score sign.
    Made with HARPA AI

  • @lalitmali07
    @lalitmali07 2 года назад +7

    Really appreciated for your dedication and effort. Thank you @Krish Naik Sir.

  • @PASALAPRAVALLIKA-hs6mt
    @PASALAPRAVALLIKA-hs6mt 6 месяцев назад

    too good to learn with ease by the help of your simple and clean explanation...thank you

  • @akhilsingh5251
    @akhilsingh5251 2 года назад +9

    After watching this playlist till now. I couldn't understand why my college teacher does not teach me like this. If my college teacher teaches me like you sir I will never bunk the classes.

    • @abd.rahaman6615
      @abd.rahaman6615 2 года назад

      Thats why They are college teacher, And Krish is a data scientist. But I didnt mean to demean a teacher.

  • @vedprasad9975
    @vedprasad9975 2 года назад +2

    HI Krish,
    This live sessions has been the most useful videos i have found in RUclips. i had viewed lot of other channels, but you the way you explain with huge patience. much appreciated. Thank you for providing this for free. thank you a ton, and you are like andrew ng for me.

  • @SBhupendraAdhikari
    @SBhupendraAdhikari 2 года назад +4

    thanks Krish concepts are very clear now...

  • @manishpatil4045
    @manishpatil4045 2 года назад +3

    Data science is easy when krish nail sir is teaching 💯❤

  • @shaolinajay9461
    @shaolinajay9461 11 месяцев назад

    Thank you sir your way of teaching is fantastic 😍

  • @anonymousmidclass
    @anonymousmidclass 2 года назад

    The way you taught stats, make me fall in love with this.

  • @dikshantakumarbharadwaj6052
    @dikshantakumarbharadwaj6052 2 года назад

    GuruJi Pranam

  • @harshitsamdhani1708
    @harshitsamdhani1708 11 месяцев назад

    thank you

  • @someshkumarmishra8279
    @someshkumarmishra8279 Год назад

    Thank you for such effort sir!!

  • @uditpujara744
    @uditpujara744 2 года назад

    really learn a lot

  • @Deoresankalp7
    @Deoresankalp7 2 месяца назад

    Sir I have request for Full Crash Course for Machine Learning that only include theory in one part and other one should be for coding problems

  • @vagheeshmk3156
    @vagheeshmk3156 Год назад +1

    #KingKrish

  • @manojrangera
    @manojrangera 2 года назад +5

    FP is type 1 error and FN is type 2 error

    • @joshivu
      @joshivu 2 года назад

      I agree.

  • @prasannasaikia9370
    @prasannasaikia9370 2 года назад +22

    Hi Krish Sir, I think you have entered the confusion matrix incorrectly. And false negative is type 2 error.

    • @talkswithRishabh
      @talkswithRishabh 2 года назад

      ++

    • @jemilmangukiya8134
      @jemilmangukiya8134 2 года назад

      If anybody know answer of this question so please help me 🙏
      1.what is use of confidence interval ?
      2.population standard deviation is given so we use z test otherwise t test, is that right

    • @ravisastry206
      @ravisastry206 2 года назад +2

      @@jemilmangukiya8134 for 1st question confidence interval is used for how "good" an estimate is and for 2nd question yeah if u don't have population standard deviation we go for T test

    • @ujjwalwadera6858
      @ujjwalwadera6858 2 года назад

      basically first we have to decide which is actual side and which is predicted one .. let's say vertical side actual and horizontal is predicted then .. 1st block will be TN , second block in horizontal will be FP , next line 1st block FN and 2nd block TP ..
      1 0
      1
      0

  • @iqrajannat
    @iqrajannat Год назад

    only 2 videos left, you are the best krish :3

  • @mallusreddy
    @mallusreddy 2 года назад +1

    Hello Krish., today I saw you video really amazing.realy the best to understand and learn for beginners. And thank you soo much for providing such a great resources... I woked in different sector and have gap. Now I am mother of twin babies. Have a long gap .I don't have time. Still making myself free to study.... But still following your videos and practicing.. Am I eligible to get job in 35+..???can you guide me.. How to follow you and how to practice in better way.. ?? Hope you will replay me...

  • @rafibasha1840
    @rafibasha1840 2 года назад +3

    8:37,i THINK Null hypothesis can never be true or accept ..we either reject Null hypothesis or fail to reject

  • @sandipansarkar9211
    @sandipansarkar9211 2 года назад

    finished watching

  • @sowndaryasubramaniyan7764
    @sowndaryasubramaniyan7764 2 года назад +5

    I have a doubt. First we used 1.96 in confidence interval formula, and then got the value of confidence interval but in the later part of video during hypothesis testing(z test) we directly used 1.96 as confidence interval without substituting in formula. Which way of calculating confidence interval is right?

    • @toyszone7749
      @toyszone7749 2 года назад +1

      1st way is for calculating "CONFIDENCE INTERVAL". The second way is for deciding H0 and H1.

  • @kunwar_08
    @kunwar_08 Год назад +1

    Till day 4 it's understandable for non maths or science background students but after dy 4 m totally confuse u can check your comment section by counting the dy 1 comments nd now on day 4 it's decreasing that means only those students are with u till now who are already knows the basic or intermediate level of stats i thought this video series will help me but I not found usefull please help

  • @prakharagarwal9448
    @prakharagarwal9448 2 года назад +6

    Hi Krish, please also upload day 4 and 5 sessions along with notes

  • @shrikantdeshmukh7951
    @shrikantdeshmukh7951 2 года назад +2

    95% is Confidence level not confidence interval you are doing good job

    • @ankitbiswas8380
      @ankitbiswas8380 2 года назад +1

      no . It is confidence interval not confidence level

    • @shrikantdeshmukh7951
      @shrikantdeshmukh7951 2 года назад +1

      @@ankitbiswas8380 please google it its or read basic book

    • @ankitbiswas8380
      @ankitbiswas8380 2 года назад

      ​@@shrikantdeshmukh7951 googled by dumb and dumber and your photo popped up .

  • @subratasaha3195
    @subratasaha3195 2 года назад +2

    Sir covarience corelation wagera kab padhenge?

  • @VinodMuda
    @VinodMuda 2 года назад

    Please add table of contents in the description

  • @kabbasoldji3816
    @kabbasoldji3816 Год назад

    i think False negative is the type 2 error , not False True negative

  • @subhodwipchatterjee4226
    @subhodwipchatterjee4226 2 года назад +2

    medication affect the intelligence, how it is shown by the null hypothesis as H0= u=100?can anyone explain me this?

  • @rafibasha1840
    @rafibasha1840 2 года назад

    1:09:50,why to subtract from 1

  • @amana.8593
    @amana.8593 11 месяцев назад

    IQ increase by 100 percent is true

  • @abhisheksinghmahra446
    @abhisheksinghmahra446 2 года назад +2

    Sir when will Be the new fsds course is coming...please make a new summer batch sir nearby april or may ...😭😭 Some of the people missed chance this time

  • @2010saumya
    @2010saumya 2 года назад +2

    Timing 33.20 you are saying sample mean can be greater than population mean how can that be then sample size has to be greater than population size which can't be true....please explain

    • @The_artt_factoryy
      @The_artt_factoryy 2 года назад +1

      Not sample size....sir is saying mean of sample may be greater than mean of population

    • @2010saumya
      @2010saumya 2 года назад

      @@The_artt_factoryy ok understood.... thanks

  • @sayanghosh7922
    @sayanghosh7922 Год назад

    I think the confusion matrix that you showed in this video is wrong.

  • @reddyprakashreddy7147
    @reddyprakashreddy7147 2 года назад

    these playlist is also for data analyst or only for data scientist

  • @shubhamsalunke616
    @shubhamsalunke616 2 года назад

    sir we took (xi-mean)/std in 2nd day and here you are taking (xbar - mean)/ std ??????? please clear my confusion

  • @kavyabhadran9911
    @kavyabhadran9911 2 года назад +1

    Can you please upload the notes ?

  • @ashusoni6448
    @ashusoni6448 2 года назад +2

    SIR when will the new batch start for full stack data science course by ineuron

    • @2010saumya
      @2010saumya 2 года назад

      I have the same question

  • @abhishekjain6694
    @abhishekjain6694 2 года назад +1

    Hi,
    In the part 1:00:20 of the video, the curve drawn is according to the 2 tail test. So shouldn't we take the value for t0.025 from the 2 tails in t table i.e. 2.492

  • @premanands1604
    @premanands1604 Год назад

    Hey Krish, is that z-test and z-score are same? as per your statement 40.32 ..i don't think so?

  • @dudeblended6495
    @dudeblended6495 2 года назад +1

    Type 1 and Type2 errors: suppose, we think about hiring process, any student selected in company with support of relatives but actually he is no able to work on that position, even he not have any skills-- Type 1 error. In case of other student , he is not selected in company but he has all skills realated to any position-Type 2 error.

  • @animesh2167
    @animesh2167 2 года назад

    Doubt: At 1:10:36 you only took Z(alpha/2) and did not multiply it with (sigma/underoot n). So Z(alpha/2) value is common for all problems/cases/scenarios?

  • @chauhanmanish682
    @chauhanmanish682 2 года назад

    18:00

  • @janauppara
    @janauppara 11 месяцев назад

    when solving last problem for 't test' in live, why did he take mean of population (mu) = 100.
    we are going for 't test' as we don't have (mu). can anyone answer if understood?

  • @vedantnimbalkar9477
    @vedantnimbalkar9477 Год назад

    Why is i alpha divided by 2 in that CI formula?

  • @cynthiac2174
    @cynthiac2174 2 года назад +1

    can someone tell me how do we conclude as medication has improved on rejecting null hypothesis? Is it because it falls > 2.05 region? how do we know null hypothesis states that it is showing the values for medication not improved on sample?

    • @sherin7444
      @sherin7444 2 года назад

      Because of of the pop mean we constructed the CI which in our case is 100.So this is considered as true.But after one sample z test we find that it is wrong since our obtained value falls outside the interval.Since the movement is to +ve side,we conclude medication has improved it

  • @Lovely_Pihu
    @Lovely_Pihu 2 года назад +1

    In confusion Matrix,
    First row: TP FN
    2nd. row: FP TN
    FP is type -I error
    FN is type -II error

    • @akashgautam1909
      @akashgautam1909 Год назад +1

      How FN can be type 2 error?
      F basically means rejecting the Ho
      And N means Ho is in fact not True.
      I think TN will be the type 2 error!

    • @LearnIndianLanguages
      @LearnIndianLanguages Год назад

      What m said is right.

    • @kabbasoldji3816
      @kabbasoldji3816 Год назад

      @@akashgautam1909 type 1 error meaning - we reject the H0 but in reality it is true
      type 2 error meaning - we accept the H0 but in reality it is false
      so let's take an example
      Our H0 = person is innocent
      but the court found him guilty, but in reality, he is innocent
      so this is a type 1 error because our H0 == reality
      but the court rejected it meaning, FALSE POSITIVE (something is false but we say it is true)
      in next case
      H0 = person is innocent
      and the court found him innocent, but in reality, he is guilty.
      so this is a type 2 error because our H0 != reality but still court accepted it
      This means its FALSE NEGATIVE (meaning, something is true but we say it is false)
      In this case, he is guilty but we say he is innocent

  • @TestProd-f1s
    @TestProd-f1s Год назад

    already you are dividing alpha/2 where alpha is 0.05 we get 0.025 in each tail why are you subtracting 1-0.025 for one tail , you can take 0.025 directly as it is for 1 tail. did you take that as we dont have values to check for 0.025 in z table? please respond

  • @arjunmr3014
    @arjunmr3014 2 года назад +2

    This live is not that good when compared with the previous 4 live sessions on stats

  • @meenakshiramakrishnan1670
    @meenakshiramakrishnan1670 Год назад

    Don't know sir

  • @voicebox4311
    @voicebox4311 9 месяцев назад

    how 3.65 is rejected it is greater than the pvalue right 3.65>1.96 right can somebody explain

    • @Awakk3n
      @Awakk3n Месяц назад

      See as 14.60>1.96 we concluded that it is beyond Confidence Interval thus rejecting the null hypothesis, similarly this time its 3.65>1.96 so we again will reject the null hypothesis.
      The question here is has the medication(in case where sample mean is 110) improved IQ or has it decreased IQ!?
      -Do let me know if you got the answer for this part, I hope you understood the above section to why we are rejecting null hypothesis( which is mu=100)

  • @vikramsinghbisen8102
    @vikramsinghbisen8102 2 года назад

    liked

  • @aftabshaaz
    @aftabshaaz 2 года назад

    Sir can you please share me these materials?

  • @milanmishra309
    @milanmishra309 11 месяцев назад

    why dont u revert on mail??

  • @nikk8604
    @nikk8604 2 года назад

    Can anybody let me know where I can find all these notes, please?

  • @siddhantgaurav7053
    @siddhantgaurav7053 2 года назад

    Sir notes toh provide krwadete Jo bhi aap likhrhe ho

  • @milanmishra309
    @milanmishra309 11 месяцев назад

    yes my IQ increased

  • @amityadav-lt5bc
    @amityadav-lt5bc Год назад

    Why the volume is so low?

  • @MrRobo930
    @MrRobo930 2 года назад

    ❤️❤️❤️🙏🙏

  • @sonalighadage6253
    @sonalighadage6253 2 года назад +2

    Outcome 2 should be Type 2 error and Outcome 3 should be Type 1 error isn't it ?
    as Type 1 => we accept the null hypothesis, in reality it is false i.e. FP and Vice versa for Type 2 (FN)
    Please correct me, if I am wrong.

    • @nitinkapoor4472
      @nitinkapoor4472 2 года назад

      You are correct , you think of it as a severity as well , Type 2 (False Negative , escape from punishment ) , Type 1 (False Positive, punished inspite being not guilty ).

    • @wonderingynow
      @wonderingynow 11 месяцев назад

      FP - Type 1 ; FN - Type 2 ..... i thinl there was a mistake when confusion matrix was drawn

  • @himanchalchandra6202
    @himanchalchandra6202 5 месяцев назад

    Why not take 36 instead of 25 to simplify things for everyone😂

  • @BrightHope-HorizonMedia
    @BrightHope-HorizonMedia 2 года назад +1

    The content is basic statistics. The title is misleading.

    • @krishnaik06
      @krishnaik06  2 года назад +11

      Please tell us what is advance the

    • @ram9208
      @ram9208 2 года назад +1

      @@krishnaik06 Hey Krish
      What do we have to find out in ATM question? CI?

    • @blast9511
      @blast9511 2 года назад

      can u tell what is advanced then for jobs and placements if u are pointing out a problem u should also suggest solution

  • @learningsinlife
    @learningsinlife 2 года назад

    my IQ increased

  • @EjazUddin786
    @EjazUddin786 8 месяцев назад

    Sir 2 years se to abi tk aap statistics cover nhi kr paye or Hindi me other further topics pr video bnane ki bat kr rhe h that not good na sir 🥲🥲

  • @azharuddinsayed
    @azharuddinsayed 10 месяцев назад

    ruclips.net/video/2VE6AZPS6bI/видео.html
    One more excellent explanation

  • @shubhendukarmakar7317
    @shubhendukarmakar7317 2 года назад

    On camera you are taking snacks does not look good. Please either abstain from doing this or stop your video. Otherwise your teaching is excellent.

  • @Navya4447
    @Navya4447 7 месяцев назад

    Thank you for your efforts ❤