FAANG Interview Question [programming] for Data Science and ML roles

Поделиться
HTML-код
  • Опубликовано: 2 дек 2024

Комментарии • 61

  • @sangram_dogra
    @sangram_dogra 3 года назад +3

    I implemented as below using recursion:
    def check(n, k):
    temp = k//10
    if temp*temp > n:
    return check(n, temp)
    elif temp*temp < n:
    for i in range(temp, k):
    if i*i == n:
    return True
    else:
    return False
    check(16, 16)

  • @awesomegaming6109
    @awesomegaming6109 3 года назад +1

    first i was going to directly look at the solution but then sir said to give it a good try and dont give up , i patiently tried and got the solution with the same basic intuition thank you sir

  • @hardikvegad3508
    @hardikvegad3508 3 года назад +1

    n = 25
    lst = []
    for i in range(1,n):
    if n % i == 0:
    lst.append(i)
    print(lst)
    for x1 in lst:
    if x1*x1 == n:
    print(f"{n} is a perfect square of {x1}")
    break
    else:
    print("Not a perfect square")

  • @RajdeepBorgohainRajdeep
    @RajdeepBorgohainRajdeep 3 года назад +6

    Sir this is an eye-opener for me. I never thought of converting any problem into an optimization problem. Sir did I missed any live session on such topics where you have discussed?

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      You can find such live sessions on numerical programming on a RUclips channel or on course website in module one under live sessions.

  • @santoshabhinav5851
    @santoshabhinav5851 3 года назад

    It's Phenomenol!! Thank you sir!

  • @asrjy
    @asrjy 3 года назад +3

    Love these videos! I have a general question. Will getting the TensorFlow developer certification help me get better oppurtunities as a fresher? I'm thinking of doing it parallelly with the course.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +1

      Yes, it will help showcase to recruiters that you understand TensorFlow well. Its acceptability in the industry is still not yet validated as the certification itself is very recent.

  • @prithvini04
    @prithvini04 3 года назад +4

    Sir pl do more this sort of videos. I'm also doing interviews prep course and finding dynamic programming a bit challenging specifically bottom up iterative approaches.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +1

      That’s why, in the interview prep course, we have 40-50 solved problems related to dynamic programming. We have another 80-100 problems to practice from. It is through persistent practice of these DP problems that you will be able to solve these problems over a period of time.

  • @smvivekhegde6488
    @smvivekhegde6488 3 года назад

    Is it possible using prime factorization? First find prime factors and check if unique prime factors are even in number..for exampe 16= 2×2x2x2,, 2 is 4 in number...100= 2x5x2x5 both 2 and 5 both are 2 in number...48= 2×2x2x2x3 here 3 is only 1 time hence not perfect square.. basically find frequency of prime factors and if every number is even in frequency distribution then it is perfect square..

  • @nileshkumar5302
    @nileshkumar5302 3 года назад +1

    n=[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]
    num=16
    for i in n:
    if i*4==num:
    print('true')
    else:
    print('FALSE', i) # here i will show the number whose sqrt is given integer by giving true and everything else false
    Is it the correct way?

  • @lusherenren4222
    @lusherenren4222 3 года назад +1

    I’ve a question, I’m in the eu. I’ve a bsc/msc in cs. By taking your course, will it help in getting into fang within the eu? Is the content/prep fang level/orientated? Thanks. Also, You’re an excellent educator.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      The hiring bar at FAANG companies is more or less similar across the world for the same roles. Our course content is at the level of Applied Scientist level roles at FAANG like companies across the world.

  • @rakshitjain5881
    @rakshitjain5881 3 года назад +1

    Can we also do our linear search till the square of the i is less or equal to our no

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      Yes, we can terminate the loop early in the linear search which will result in a time complexity of O(sqrt(n)) which is worse than binary search.

    • @rakshitjain5881
      @rakshitjain5881 3 года назад

      Yes but if interviewer asked for optimized solution of linear search we can go for this also

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      Interviewers will ask for the most optimal solution you can build. Obviously improving the linear search to O(sqrt(n)) is a good starting point. But using binary search based method is better.

  • @bhavanakaturi916
    @bhavanakaturi916 3 года назад

    Count the number of distinct prime factors for that number other than 1. If all the counts are even then it is a perfect sqaure else it is not a perfect square
    E.g: 16->2,2,2,2 -number of 2's is 4(even)
    Then it is a perfect square
    E.g 18->2,3,3-number of 2 is even but 3 is odd ,hence not a perfect square
    Is this approach valid?

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      This would work. But, how would you find the prime factorisation? What is the time complexity to find that?

  • @shubhamchoudhary5461
    @shubhamchoudhary5461 3 года назад

    i'm beginer but i can try...
    n = int(input('enter num'))
    if type(n**0.5) == int :
    return 'perfect_square'
    else:
    return 'not perfect_square'

  • @pranayreddygujjula1551
    @pranayreddygujjula1551 3 года назад

    n=256
    flag = False
    for i in range(1,n//2):
    if(i**2 == n):
    print('True')
    flag = True
    break
    if(flag == False):
    print('False')
    Hi Team,
    Please give feedback for the above code

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      This is exactly like the linear search, right. You can always test it fairly easily against the test cases that we discussed in the above video.

  • @trilokchandra7469
    @trilokchandra7469 3 года назад +1

    Sir can you please tell what is the avg package of students who are placed after completing AAIC???

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      This depends a lot on the prior experience of the candidate and how well they’ve learnt from the course. The course content itself is sufficient to crack most of the roles, including those at the top best companies. Most freshers or those with less than two years experience get in the 6-12L range with a median of 8L. For experienced professionals, the range of compensation increment can be anywhere from 30 to 250% based on how well they picked up the foundational skills.

    • @himalayasaikia5762
      @himalayasaikia5762 3 года назад

      @@AppliedAICourse when you say experience people...does people from non data science experience also get 30-250% hike

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      Yes, work experience. Not necessarily in ML/DS.

    • @trilokchandra7469
      @trilokchandra7469 3 года назад +1

      @@AppliedAICourse sir for 1year experienced in web app developmenter will the placement is guaranteed after completion of course
      I mean a full time job placement or an internship

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      It will be a full-time role. In the worst case, you might end up joining as an intern and then get converted to a full time. Very often, people with experience do not join as interns unless they have career gaps.

  • @ankesh23569
    @ankesh23569 3 года назад

    Sir can please tell me which metric we should use in Recommendation systems

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      There are many possible metrics that can be used for recommendation system problems. A popular option is mean average precision at K. You can read more about it and various kaggle competitions that use it at: datascience.stackexchange.com/questions/82463/metric-mapk-for-what

    • @ankesh23569
      @ankesh23569 3 года назад

      @@AppliedAICourse Ok sir I'll check it out and Thank you very much for your great concern regarding this RUclips community.

  • @srijeetful
    @srijeetful 3 года назад

    Very nice video sir , I have one question like ""how much of cloud/ devOps/ API technologies as an (1)ML engineer (2)Data Scientist and (3) Applied Scientist, you are supposed to know in a good tech company (like FAANG, Microsoft, Adobe ) ?"

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +1

      Most of these companies expect strong core skills and programming knowledge. Even if you don’t know Cloud/DevOps/APIs, they will hire you if you have strong mathematical, programmatic and applied knowledge of data science and machine learning.

    • @srijeetful
      @srijeetful 3 года назад

      @@AppliedAICourse Thanks a lot Team for the fast response !!! It is very helpful

  • @venkateshlotlikar1621
    @venkateshlotlikar1621 3 года назад

    def perfsq(n):
    if n < 0:
    return False
    isperf = False
    for i in range(n+1):
    if i**2 == n:
    isperf = True
    break
    return True if isperf else False

    n = int(input())
    print(perfsq(n))
    Can you please provide feedback on this code?

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +1

      This is exactly like the linear search that we described in the video. Please try and implement binary search on your own as it has a better time complexity.

  • @shortsbytuffy_shero8430
    @shortsbytuffy_shero8430 3 года назад +1

    Sir I will be in 1st year of college this year as I got decent marks in jee mains and 100 percent sure of getting MNNIT ALLAHABAD cse .
    I am learning c++ as a beginner what should I do to improve my coding skills , I want to be a game developer??

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +1

      Focus on being a good programmer and learn data structures and algorithms. Write as much code as possible in your first year and become really good at coding and related problem-solving using data structures and algorithms. That will build a very strong foundation to whatever career you want to pursue as your interests will evolve and change over the next four years.

    • @shortsbytuffy_shero8430
      @shortsbytuffy_shero8430 3 года назад +1

      @@AppliedAICourse sir is there any course of yours for competitive coding??

    • @shivamjalotra7919
      @shivamjalotra7919 3 года назад

      Game Developer + 1st Year, Hmmm Makes Sense.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      interviewprep.AppliedRoots.com

    • @shortsbytuffy_shero8430
      @shortsbytuffy_shero8430 3 года назад

      @@shivamjalotra7919 will do it ,cause I dropped a year even after getting nit jamshedpur civil just for cse in some decent nit computer science because I don't want to crush my dream..and yeah I took the decision of drop in December 2020 after counseling...

  • @saaiswetha7056
    @saaiswetha7056 3 года назад

    these coding questions can be solved using only python? or there are other language preferences? Which language is usually preferred for coding rounds in Data science roles?

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      Most top product-based companies would be okay with you implementing in any of the major programming languages.

    • @saaiswetha7056
      @saaiswetha7056 3 года назад

      @@AppliedAICourse does this apply irrespective of the role we're applying for?

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      In most cases, you are given flexibility to choose a language.

    • @saaiswetha7056
      @saaiswetha7056 3 года назад

      @@AppliedAICourse thanks!!

  • @shivamjalotra7919
    @shivamjalotra7919 3 года назад

    1. The gradient descent method is also called Newton's Method of finding roots.
    cp-algorithms.com/num_methods/roots_newton.html
    2. There is one more O(sqrt(n)) solution based on prime factorisation of n.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +2

      Yes, there are many such good methods to solve this problem. Note that O(sqrt(n)) is worse than O(log(n)). The Newton’s method of finding roots is often taught at undergraduate level in the first year coursework.

    • @MahJohn
      @MahJohn 3 года назад

      Actually, the gradient descent isn't called Newton method. It's however, in another way called steepest-descent or cauchy's method. It was first proposed by a great Mathematician named Augustin Lious Cauchy around 1847.

  • @tejaskumar453
    @tejaskumar453 3 года назад

    why can not we use simple math-
    res = ((x)**(1/2))%1
    if res>0:
    not perfect square

    • @AppliedAICourse
      @AppliedAICourse  3 года назад +2

      x**0.5 is an indirect way to computing the sqrt() function, which we suggested should be avoided in the problem definition.

    • @abhishekpodder7594
      @abhishekpodder7594 3 года назад

      @@AppliedAICourse But its not explicitly mentioned. That power of 0.5 has to be avoided.

    • @AppliedAICourse
      @AppliedAICourse  3 года назад

      But in a typical interview, the interviewer will quickly point out and say that it is equivalent to using SQRT function. He would ask you to rewrite it without using it.

  • @user-yn8bh
    @user-yn8bh 3 года назад

    Is this code is more optimized then binary search solution , check it.
    And also tell , is this code can be more optimized with linear search only ?
    n=[4,9,8,7,36,35,49,0,10001,10000,100000]
    a = max(n)
    lst=[]
    x=0
    y=1
    while x