Step by step roadmap to learn data science in 6 months | Complete data science roadmap

do you want to learn data science but not sure  where to start because on internet there is so   much overwhelming information or you are someone  who is not from a technical field and you have   doubts that you don't know programming or any  technical details and can you learn data science   in this video i am going to show you step-by-step  approach of learning data science mostly by using   online free resources in 6 month time period  i'm assuming you do four hour dedicated study   with focus every single day and by doing that you  can learn data science now data science learning   is an ongoing journey i'm still learning today the  six month will give you a very solid base so that   you can start doing data science project so many  people have doubts that they don't have technical   background and can they learn data science  i have a physiotherapist friend who wanted   to jump into data science industry he did some  courses and he's working as a data scientist   so if a physiotherapist person who is from a  totally irrelevant background can do it you can   do it also i know so many people and information  is available online for free so all you need is a   laptop internet and a solid willpower believe it  or not so many data scientists and data analysts   use excel in the their day-to-day life so excel  is the first step in your data science journey   you need to learn certain things such as  macros formulas how to do filtering vba   pivot table vlookup etc in excel excel is a  very basic tool that allows you to data science   data science after all is a process of drawing  insights from your data and when we did not have   big data or programming languages people  were using excel to do the data science your first excel data science project could be  managing your own personal finances on google docs   there is a template gallery for personal  finances so if you click on this link   you will find these templates where you can  track your monthly incomes and expenses so   if you click on lesson monthly budget you  will find this excel file where you can enter   all your monthly expenses and then it  will tell you how much money is left   after doing all these expenses so you're doing  subtraction basically and you can also maintain   detailed transaction now in this excel file if you  look at any of these cells you see the formula at   the top at the top so these are the formulas  that you want to learn so that you can create   nice visualizations like this because  as a data scientist or data analyst   you will be involved in creating these kind of  visualization charts and so on and working on   your own personal finances help you with that  you can also do annual budget tracking where in   one sheet you are maintaining all your monthly  expenses on children that education and so on   in the other sheet you have all your income and  then you get a nice summary so if you look at   these charts right see they have these formulas  so these are the formulas you want to learn   what you want to do is you want to create your own  excel file that can maintain your monthly expenses   and uh income and in the end you can plot the  trends in which month you are having the maximum   expense what is your trend of income overall  month to month is it increasing or not so all   of this you will be able to do in excel and in the  end you are producing a tool which is helping you   so this gives you a mental reward and you can  continue on your learning journey with passion   there is this another website which has a lot of  templates so i i downloaded this family budget   planner you will see nice planner like this where  you can plot even these kind of bar charts see if   you look at this bar chart there is some formula  being used and this is the formula that you want   to learn so if you fill all these numbers here  you will see these charts being updated so your   goal is to take inspiration from these files and  you create your own uh budget tracker that will   be the best beginner excel project that you can  work on there is a youtube channel called chandu   and i suggest that you learn excel things from  here this guy is great he explains excel really   well he also has this website called chandu.org  so if you are looking for any excel related help   refer to this website and a youtube channel while  you're learning excel you also want to tackle   statistics because statistics is used everywhere  so the first resource i suggest is can academy's   basic course so if you look at this course it  has a very nice explanation using visualization   it also has exercises so make sure you complete  this course and you will learn basic things about   what is normal distribution the inferential  statistics the descriptive statistics   probability the sample distribution uh summarizing  quantitative data and so on these are the   concepts you want to start with and as you do more  projects you can learn advanced topics there is   also a youtube channel for khan academy so if you  are if you like watching youtube you can watch uh   the videos here as well but if you want that  course you will get a benefit of doing exercises   so i suggest doing that versus a youtube channel  there is another youtube channel by martin states   lecture this is also another great platform  where you can learn things like different   kind of charts histogram and density plots box  plot is used for outlier removal so you want to   learn all these basic things the other channel  that i suggest is stat quest with joe stremmer   this person explains complex concepts in a very  simple manner using visualizations so if you have   any doubt on statistics any topic you're not sure  about please refer to this channel you will spend   first two weeks in learning excellent statistics  after that you should move on to the next topic   you don't want to get stuck into one or two  different topics you want to do parallel learning   so don't spend like three months just learning  excellent statistics two week time period is   enough every day you're putting four hours if  you are putting less time then spend little   but don't spend like two or three months  just learning excellent statistics   because then it will kind of demotivate you so  whatever you can learn in this much time period   you learn it and then you move on to the next  topic the next topic is a programming language   you want to choose between python and r these are  the two most popular programming languages in the   field of data science if you are not sure  always go with python it is very versatile   and it has a rich set of libraries now many  people get afraid of programming language they   are like ooh programming i cannot do it but  trust me it's not that hard python especially   is english like it is so easy to learn even  a high school student can learn it easily   nowadays in the high school they teach it and  when i was going to a nearby public library   to teach python as a volunteer there was an  eight-year-old kid who was learning python   so even small kids are learning it so you guys can  do it you need just little bit of patience and i   have this complete tutorial uh playlist on python  and you want to follow first 16 videos in it   you don't need any prior knowledge of programming  to follow this tutorial i start with the   installation then go over basic data types such as  number string and list and then we discuss about   control flow statements if for loops and such then  we cover functions dictionaries and tuples modules   working with json etc you need to just do first  16 tutorials you can do advanced tutorials but   you don't want to spend crazy amount of time just  looking at python because python itself is so huge   that's why i'm saying only 16 videos  and then you move on to the next topic   when you're doing this video for  example reading and writing files   make sure you check video description because  that has an exercise link so if i click on this   link here i will find an exercise so you watch  a video learn things and you work on exercises   it is very important to consolidate your learning  after you do your own solution you can click on   the solution link to verify your answer with my  answer do not click on solution link directly   otherwise it will download the corona virus to  your computer and your computer will get a fever   so you get the point right you want to work on  it on your own first then verify with my answer   now i have designed this tutorial playlist without  using any jargons it is perfect for very average   or below average person if you decide to learn r  instead of python then i suggest marin's lectures   so that's another youtube channel so you can  just follow this tutorial playlist and learn r   you want to spend next one month which is week  number five to eight in learning numpy pandas and   data visualization library these tools allow  you to do data cleaning and data exploration   data scientists spend 70 percent of their time  in doing data cleaning and exploration the model   building part is only 20 or 30 percent so it's  very important that you have a good grasp on numpy   pandas and visualization library which could be  either matplotlib or c bond data cleaning process   allows you to clean up your messy data and put it  in a format such that it can be used by your model   i have this great visual from quora which i just  absolutely love and it shows you know how the man   is stuck in those wires and by doing numpy pandas  data cleaning all you're doing is putting those   wires on nice reels so that you can use that clean  data for later processing i have all necessary   links required to learn all the tools which i just  mentioned in your second month of data science   learning the first one is numpy tutorials i have  this four simple tutorials it shouldn't take you   more than two three hours to go through it and  there are jupiter notebooks and the code available   so you can just practice while you're watching  these videos numpy is much better than python   plane list because python plane list is slow  it's not optimal in terms of memory utilization   numpy is ideal for storing huge arrays when you're  doing data science you're often dealing with huge   arrays with lot of numbers numpy internal is  implemented in c plus person fortran and it is   super fast pandas is built on top of numpy  and the main object in pandas is data frame   data frame allows you to store tabular data in  memory and after that you can perform sql type of   operation on that data frame so just imagine  you are getting all the benefits of a database   in memory and it is super fast when you are  following this tutorial playlist for pandas   you will start with data frame basics then you  will learn how to read and write data from you   know from csv in excel files then the next two  tutorials are on handling missing data in data   science data scientists spends a lot of time  actually in handling the missing values you can   also do group by concat merge etc so you need to  follow the tutorial up till tutorial number nine my binders tutorials have received a  great response from youtube you can   see that from the views and from the likes and  these tutorials are perfect for any person who   is just beginning the journey i don't use  any jargons and try to explain things in a   simple way also in the video description you  will find my jupyter notebooks the next one   is matplotlib which is a visualization library for  that you need to just follow these seven tutorials   where you'll go over different type of charts  such as bar chart histogram pie chart and so on   matplotlib is very easy to use compact library  but sometimes if you want to do some advanced   stuff you can use c bond now i have started c bond  series with the help of vedant recently right now   the playlist has on only one video but we are  uploading the videos fast so it's likely that   when you're watching this video in future  will have probably the complete series   after you learn the basic syntax on numpy pandas  matplotlib your next thing would be to try   exploratory data analysis or eda on real data  sets from kegel so here i have kegel notebooks   so if you look at this url gaggle.com notebooks  and you will find so many notebooks here for   example this is an eda notebook so when you  look at this notebook you will get an idea on   what type of explorative data analysis you can do  using pandas and metaplot labency born and so on   giggle also has a lot of data set so if you go  to kegel.com datasets you will find this data set   on which you can perform ada for example if you  are a food lover there is an indian food data set   so just download this and open a jupyter  notebook and start doing your small projects   so after following tutorials when you do small  projects you consolidate your understanding   and you develop solid data analysis skills  which will be useful in data science process   once you know a little bit about exploratory  data analysis now it's a time to jump onto   machine learning so you want to spend your next  one month in learning machine learning now if   you're targeting a career of data analyst machine  learning learning knowledge is not that much   needed of course if you have the knowledge is good  but it's not needed per se as per the job role   but if you're targeting data scientist role  then machine learning knowledge is a must   now when you will learn machine learning you will  find this difficulty that you will come across lot   of math and statistics concept now we only spend  a week or two in learning statistics so now what   you want to do is when you are doing machine  learning as you go on those tutorials you should   continuously learn new mathematics and statistics  concepts again you can use khan academy different   youtube channels which i mentioned there is an  another book called think in stats which i'm   going to link again in the video description so  you can use all these awesome resources if you're   not sure about mathematics math is fun dot com is  a very basic very fun uh website you know they say   math is fun and they actually make math learning  fun another youtube channel is 3b1b where uh that   person is teaching math in a really very intuitive  way using visualization so now your process is   you're learning machine learning you come across  let's say logistic regression you don't know what   it is now you want to stop there you want to go  to your math and statistics so you kind of get   an idea that it's not like you're picking up one  topic and just learning everything and not looking   at it again you have to always revisit uh those  same topics uh to make your concepts clear i have   a complete machine learning tutorial playlist for  a beginner with all the concepts if you click on   this link you will find this particular playlist  where we begin with the basics of machine learning   then we cover linear regression gradient descent  dummy variable hot encoding etc then we touch   base on classification techniques such as logistic  regression decision tree support vector machine   you need to know regression classification  and clustering techniques in your basic   machine learning course and i have all  of that covered here after that you need   to know hyper parameter tuning how do you  select a based model for a given problem   once you have completed 16 tutorials you can  now embark on a end-to-end data science project   here i have this eight part tutorial series for  building a website or an end-to-end project where   i cover web development model building python  flash server deploying it to amazon aws and so on   if you look at the videos of that project in the  video description you will find a link to a code   repository and if you go to that repository  you'll find a complete code for that project   here we built this website where you can supply  different parameters for an apartment and when   you click on estimate the machine learning model  will estimate the price of that particular home   in model directory we built the model did fine  tuning there is a server client and so on this   is a complete uh project which will be extremely  beneficial to you also if you're looking at any   of my machine learning tutorials i always  include exercise at the end so you can open   this particular link where you see the notebook  which i had covered for that particular video and   then in the end you will find exercise description  there is also exercise solution so once you try   to solve the problem on your own you can tally  your answer with my answer by looking at this   when you search for machine learning tutorial  python you'll find my videos in first five or   six search results and these videos are very very  basic way i explain things in easy to understand   manner so the the material is basically perfect  for the beginners as i have outlined here you want   to spend one month in machine learning and doing  at least one project i have more projects too   like i have image classification but probably  you should do that later so your goal is to do   only one project and basic uh study of machine  learning and then move on to the next topic   after you have understanding of machine learning  you want to dedicate it next one month in learning   deep learning deep learning is all about neural  network it starts with artificial neural network   then we move on to convolutional neural network  which is used mainly for image processing   then we move into rnn or recurrent neural network  that is mainly used for nlp week 13 to 16 is dub   deep learning as outlined in this plan if you open  this youtube playlist this is the one for covering   all deep learning concepts as of this recording  i'm still working on it and i have tutorials still   convolutional neural network i'm yet to cover rnn  but it's possible when you're watching this video   you'll have all the videos available now say  you are watching a video on derivatives here   i try to explain the theory behind derivatives  then i do some coding and in the end you have   an exercise so if you open the exercise it will  have exercise description along with the solution   so i i keep on repeating about this exercise  because learning is not just about watching   video it is you doing coding along with the video  and then working on exercise very important as a   data scientist it's not that you will be dealing  with excel or csv files you will most likely be   dealing with databases and there are two type  of databases sql database and nosql so for sql   mysql is the most popular database there is oracle  and some microsoft relational databases as well   and when it comes to nosql there are databases  like mongodb so you need to have a good   understanding of these databases because if you  are working as a data analyst or data scientist   your job will be to connect to those databases  and pull the information that you need   for doing data science so if it is mysql database  you need to have understanding on sql queries   if it is nosql you need to know certain nosql  fundamentals so that you can retrieve the data   that you need for the data science process so here  is one month schedule to learn the databases the   first playlist i recommend for sql is cuda venkat  which is a popular playlist on youtube here you   need to follow the video still 16 or 15. you can  follow a rest of the tutorials but the knowledge   that is needed to start your data science  project you will get by doing the first 16 videos   there is a khan academy sql course as well  so i suggest that you do this course first   and then you watch the cuda banker playlist this  course has very good explanation of sql concepts   relational databases joins and queries  and where clause select clause and so on   you need all of these when you are doing  data retrieval in your data science process   so follow these two resources to learn about  sql databases for no sql there is this coursera   course it's a three-week course but you don't  need to follow all the materials just follow   first few videos just to get an understanding  of mongodb coursera i have heard that you can   enroll it for free and only if you want to get a  certificate you'll have to pay pay the fees but   you can whichever country you are in you can just  check the course there are terms and conditions   now you want to spend your last month in your  journey of learning data science in learning about   bi tools so the popular bi tools are tableau  power bi and click sense and these tools allows   you to connect with your data source it could be a  database your csv file whatever and you can build   nice dashboards or interactive visualization these  tools are so powerful that you can almost build   a mobile application without doing coding so  if you follow my sales insight project series   for power bi in that project series we did this  project in a corporate style where you take   sales inside data and you build not only powerful  virtualization you can almost export those as a   mobile app so the chief operating officer of that  company or a sales director of that company can   have a interactive mobile app and they can look  at all the different visualization these tools are   so much powerful especially if you're targeting  data analyst career you need to have knowledge   on these tools here is the curriculum from week 21  to 24 for learning bi tools for tableau abhishek   agrawal's youtube playlist is pretty good you have  so many tutorials on tableau you don't need to go   through all of them but just get through basics  basic post 10 15 tutorials and get an idea we also   have another playlist by apartheid w consultancy  this tutorial playlist is also good and then   there is this code basic sales insights playlist  i am planning to do another tableau project so   if that project is done in the future i will  mention the link here in this same github md   page by the way this page is available in the  video description so don't forget to check it out   so after this six months you would have gained  all the necessary skills needed for becoming   data scientist or a data analyst you need to also  learn some soft skills because there is so much   things to learn you must be overwhelmed you'll  be like dude when am i going to learn all these   things just too much for this reason you need to  know how to learn things effectively for that also   i have a video so before you start your learning  process i suggest that you watch this video first   because you want to spend minimum amount of time  and you want to get maximum output and this video   will teach you how you can do that extremely  effective focused learning another great idea   of making learning effective is to do group study  if you do it on your own it's it's possible that   you will go off the track you will be demotivated  but if you make a group let's say you are studying   in a college you make a group with your  college friends if you are in at workplace   you might help similar people who want to learn  data science so you make a group if you don't know   where to get that group buddy i have a discord  channel and under that discord server there is   a channel called you know like a partner finder  and in that channel people are posting about   group studies so you can make a buddy uh by  following that channel my discord server the   link of the discord server is below in the  video description it is free you can join it   and make use of that we have pair programming  channels as well so you know two people can do   pair programming uh using this code all right  now it's a time for the giveaway yes we are   doing a giveaway for this video so what i want  you to do is a post in a video comment one idea   if you have three ideas put three different  comments and the idea is you want to   describe how machine learning deep learning and  data science is going to change the future so this   is more futuristic for example you can post in  a comment saying that we will have autonomous uh   cars right now autonomous cars is something we  all know uh but other than autonomous cars if you   think about any other areas where deep learning  machine learning will have a very big impact   then i want you to post that idea again if you  have three ideas post three different comments   and this giveaway is open for three days so  we will collect all the comments received in   next three days after posting this video so  when the video is live premiering after that   we will have three days to collect all the  comments and the first three top comments will   receive code basics t-shirt so these  are the t-shirt nicely painted t-shirts   and you will receive those t-shirts as a  gift so just use your creativity and post   an idea about how data science and deep learning  and machine learning is going to change the future   i hope you like this video or trust me learning  data science is not that hard all you need is a   discipline and follow this step-by-step roadmap i  have mentioned i have given a link of this roadmap   in the video description below so don't forget  to check it out and do not have so many doubts   in your brain just start it today okay do not  wait the world is evolving so fast and there   are so many amazing things happening and there are  so many opportunities you can build an excellent   carrier on a lot of money and live a life full  of thrill and passion so i wish you all the best

Channel: codebasics

Views: 692,307

Keywords: data science roadmap, learn data science smartly, learn data science step by step, data scientist learning path, how to learn data science smartly, data science roadmap 2021, complete data science roadmap, data science roadmap 2022, road map to data scientist, data science full course, data science for beginners, how to start data science learning, how to learn data science, how to become data scientist, how to start data science learning for beginners, data science tutorial

Id: H4YcqULY1-Q

Length: 30min 42sec (1842 seconds)

Published: Sat Oct 17 2020