How I'd Learn Data Science In 2024 (If I Could Restart) - The Ultimate Roadmap

Summary

This video script serves as an ultimate roadmap for aspiring data scientists, detailing a comprehensive learning path for 2024. It emphasizes the importance of understanding data science basics, acquiring key skills in Python and mathematics, and applying them through project-based learning. The guide also advises on job applications, further specialization, and staying updated with industry trends. It encourages building a strong foundation, exploring various data science domains, and networking within the community to excel in the field.

Takeaways

• 🧭 Start with self-assessment: Understand what data science is, the skills required, and if it aligns with your interests before diving in.
• 💡 Research and observe: Spend time researching data science, looking at its applications in various industries, and understanding a typical data scientist's workday.
• 🛠️ Programming first: Prioritize learning programming over math, starting with Python due to its prevalence in job postings and community support.
• 🐍 Choose Python: Focus on Python as it is sought after by 60% of job postings, providing a wider range of opportunities compared to R.
• 💻 IDE selection: Use a popular and user-friendly IDE like Visual Studio Code (VS Code) for writing Python code.
• 📚 Learn Python basics: Grasp fundamental Python concepts such as data types, variables, lists, dictionaries, and pandas data frames.
• 📊 Visualization tools: Learn to use visualization libraries like Matplotlib, Seaborn, or Plotly to create charts and graphs for data representation.
• 🔁 Project-based learning: Apply newly learned skills through projects to solidify understanding and make learning more practical and memorable.
• 🔍 Specialize wisely: After mastering the basics, consider specializing in areas like machine learning, NLP, or computer vision based on job market demand and personal interest.
• 🔗 Networking and community: Engage with data science communities and networks, both online and in-person, for support, collaboration, and knowledge sharing.

Q & A

What is the main goal of the video?

The main goal of the video is to provide a comprehensive blueprint for becoming a data scientist in 2024, including the most effective learning path, projects to work on, resources for learning, and ways to stand out among other data scientists.

Why does the video emphasize starting with programming rather than math?

The video emphasizes starting with programming because it forms the foundation for implementing mathematical concepts in data science. Even if one is not great at math, they likely have some familiarity with it from high school, whereas programming can feel like a new world and is essential for coding in data science.

What programming language does the video recommend learning first for data science?

The video recommends learning Python first because it is more widely requested in job postings and offers more employment opportunities compared to R.

What are the six basic concepts in Python that the video suggests learning?

The six basic concepts in Python suggested by the video are data types, variable assignment, lists, dictionaries, Pandas data frames, and basic control flows (IF statements, for loops, while loops).

Why is project-based learning emphasized in the video?

Project-based learning is emphasized because it helps solidify the knowledge gained from learning programming and mathematical concepts, making it easier to remember and apply these skills in real-life situations.

What are three beginner-friendly projects suggested in the video to solidify Python skills?

The three beginner-friendly projects suggested in the video are creating a simple contact book application, building an inventory management system, and writing a function to analyze an Excel file and return basic descriptive statistics.

How does the video advise approaching the job market while still learning?

The video advises gently prodding the job market by applying for entry-level roles and internships without full customization of the application. This helps to understand the job market's demands and to test and improve one's CV.

What are some areas of specialization mentioned in the video for data scientists?

The video mentions natural language processing, anomaly detection, predictive modeling, recommendation algorithms, marketing mix modeling, computer vision, and general machine learning as areas of specialization for data scientists.

Why is SQL considered a valuable skill for data scientists according to the video?

SQL is considered valuable for data scientists because it is used for querying and creating databases, which is a skill primarily used by data engineers and data analysts but is still beneficial for data scientists to understand and utilize.

What is the video's stance on the importance of having a digital footprint for data scientists?

The video suggests that having a digital footprint, such as posting about your journey on LinkedIn and Twitter, can help data scientists stand out as it shows their progression and engagement with the field, which can be appealing to potential employers.

Outlines

00:00

🚀 Introduction to Becoming a Data Scientist

The paragraph introduces the concept of becoming a data scientist and outlines the roadmap to achieve this goal. It emphasizes the importance of understanding what data science is, the skills required, and whether it aligns with one's interests. The speaker encourages research, looking into industries that utilize data science, and understanding the typical workday of a data scientist. The focus is on answering three key questions before diving into the field: What is data science? What skills are needed? And is it the right career choice? The paragraph also stresses the importance of learning programming, specifically Python, as a foundational skill for data scientists.

05:01

💻 Building a Solid Foundation in Programming

This section delves into the importance of programming, particularly Python, as the cornerstone of a data scientist's skill set. It advises against spreading efforts across multiple programming languages and instead recommends focusing on Python due to its prevalence in job postings. The paragraph introduces the concept of an IDE (Integrated Development Environment) and suggests Visual Studio Code as a user-friendly option. It then lists six fundamental Python concepts essential for data science: data types, variable assignment, lists, dictionaries, control flows, and functions. The speaker also advocates for project-based learning to solidify these concepts, suggesting beginner-friendly projects like a contact book application, an inventory management system, and a function to analyze Excel files.

10:01

📈 Advancing with Pandas and Data Visualization

The paragraph discusses the transition from basic Python skills to more advanced data manipulation with Pandas, focusing on data frames. It simplifies the concept of data frames as 'fancy tables' akin to those used in Excel. The speaker introduces three additional fundamental concepts: basic control flows, data visualization libraries like Matplotlib and Plotly, and functions. The importance of project-based learning is reiterated with the suggestion to apply these newly learned skills in practical projects. The paragraph concludes by encouraging self-congratulation for overcoming the initial challenges of learning to code, emphasizing that the acquired knowledge is useless unless applied through projects.

15:02

🔢 Integrating Mathematics into Data Science

This section introduces the integration of mathematical concepts into the data science learning journey. It emphasizes the importance of understanding statistical concepts, linear algebra, basic calculus and trigonometry, and probability. The paragraph suggests focusing on fundamental mathematical concepts that are relevant to data science and provides resources for learning, including YouTube channels and random Google articles. It also suggests project-based learning to solidify mathematical knowledge, proposing projects like calculating moving averages, implementing statistical functions, and exploring libraries like NumPy and SciPy.

20:02

🌐 Specializing and Enhancing Employability

The paragraph discusses the importance of specializing in certain areas of data science to enhance employability. It advises having a solid understanding of basic data pre-processing, feature engineering, and supervised learning. It then suggests exploring areas like natural language processing, anomaly detection, and machine learning for deeper knowledge. The speaker recommends starting to apply for jobs early in the learning process to test the job market and refine the CV. The paragraph also introduces additional skills like SQL, data visualization with Tableau, and the importance of community and networking. It concludes by encouraging the pursuit of a data science community for support and knowledge sharing.

25:03

📊 Mastering Data Presentation and Community Building

This section focuses on the importance of data presentation skills, particularly with Tableau, to create appealing dashboards that can help in job applications. It also touches on the value of community building and networking for moral support and problem-solving. The speaker suggests forming a community for dedicated learners and professionals to accelerate their data science journey through study sessions, calls, and mentorship. The paragraph concludes by encouraging the pursuit of additional skills like working with APIs, using GitHub, and learning streamlit, as well as maintaining a digital footprint on professional platforms to showcase one's journey and expertise.

📈 Staying Cutting-Edge in Data Science

The final paragraph emphasizes the importance of staying up-to-date with the latest trends in the rapidly evolving field of data science. It suggests following data science professionals on platforms like Medium, YouTube, Twitter, and LinkedIn for insights, tutorials, and discussions on new techniques. The speaker also encourages subscribing to newsletters and engaging with the data science community to continue learning and growing in the field. The paragraph concludes with a reassurance that the roadmap provided is comprehensive and encourages subscribers to access written resources for further guidance.

Keywords

💡Data Science

Data Science is an interdisciplinary field that uses scientific methods, processes, algorithms, and systems to extract knowledge and insights from structured and unstructured data. In the video, it is the central theme, with the speaker providing a roadmap for aspiring data scientists. The video aims to clarify what data science entails and the skills required to excel in this field.

💡Python

Python is a high-level programming language that is widely used in data science for its simplicity and the powerful libraries it supports for data analysis and machine learning. The video emphasizes learning Python as a foundational skill for data scientists, noting that many job postings explicitly ask for Python proficiency.

💡IDE (Integrated Development Environment)

An Integrated Development Environment is a software application that provides comprehensive facilities to computer programmers for software development. In the context of the video, the speaker recommends using an IDE like VS Code for writing Python code, highlighting its popularity and utility among developers.

💡Pandas

Pandas is a Python library providing high-performance, easy-to-use data structures, and data analysis tools. The video mentions Pandas as a crucial library for data manipulation, particularly for working with data frames, which are akin to tables and are fundamental to handling data in Python.

💡Mathematics

Mathematics, particularly statistics, linear algebra, calculus, and probability, forms the backbone of many data science algorithms and techniques. The video underscores the importance of having a strong grasp of mathematical concepts to interpret and apply data science methods effectively.

💡Project-Based Learning

Project-Based Learning is an educational method where students gain knowledge and skills by working for an extended period on investigations, research, and projects. The video advocates for this approach in learning data science, suggesting that applying newly learned skills to real-world projects solidifies understanding and makes the learning process more meaningful.

💡Statistical Concepts

Statistical Concepts such as mean, median, mode, variance, and standard deviation are essential in data analysis for summarizing and describing data. The video includes these concepts as part of the mathematical foundation necessary for data scientists to analyze and interpret data accurately.

💡Linear Algebra

Linear Algebra is a branch of mathematics that deals with linear equations, vectors, matrices, and linear transformations. The video mentions that understanding linear algebra is crucial for data scientists as it provides the framework for many operations and algorithms in the field.

💡Visualization

Visualization refers to the presentation of data in a graphical or visual format, making it easier to understand and interpret. The video suggests learning visualization libraries like Matplotlib and Plotly to create charts and graphs that can help in presenting data insights more effectively.

💡SQL

SQL (Structured Query Language) is a domain-specific language used in programming and software engineering, as a database query tool, especially the requested query for online transaction processing, but is also used for database management, and data analysis. The video highlights SQL as an additional skill that data scientists should learn, as it is valuable for querying and managing databases.

💡Machine Learning

Machine Learning is a subset of artificial intelligence that provides systems the ability to automatically learn and improve from experience without being explicitly programmed. The video touches on machine learning as an area where data scientists can specialize, and it is a field that is constantly evolving with new techniques and applications.

Highlights

By the end of this video, you will have a comprehensive blueprint to become a data scientist in 2024.

The video provides a detailed roadmap, including the most effective self-taught route, project ideas, and resources for learning data science.

It emphasizes the importance of answering three fundamental questions before starting to study data science.

Research is suggested to understand data science, its applications in various industries, and the typical workday of a data scientist.

Programming, particularly Python, is recommended as the first skill to learn for data science, ahead of mathematics.

Python is chosen over R for its broader demand in job postings and less competition for those positions that do require R.

Visual Studio Code (VS Code) is recommended as an Integrated Development Environment (IDE) for writing Python code.

Six fundamental Python basics are outlined as essential for data science, including data types, variables, lists, dictionaries, and control flows.

Pandas and data frames are introduced as important for working with data tables in Python.

Project-based learning is stressed as a method to solidify knowledge, with three beginner-friendly projects suggested.

The video discusses the benefits of structured learning through courses or boot camps over self-teaching via YouTube.

DataCamp is recommended for learning Python, SQL, and other data science skills.

Mathematics is introduced as a concurrent learning path, focusing on statistics, linear algebra, calculus, and probability.

Applying for jobs with a basic understanding of math and programming is advised, even before mastering all skills.

The importance of tailoring a CV to highlight data science skills and projects is emphasized for job applications.

Specialization areas in data science such as NLP, anomaly detection, and machine learning are suggested for deeper knowledge.

SQL is highlighted as a valuable skill for data scientists, with a focus on querying, database creation, and working with relational tables.

Visualization skills, particularly with Tableau, are recommended to present data science findings effectively.

The video concludes with advice on staying up-to-date with the latest trends in data science through platforms like Medium, YouTube, and following industry leaders.

Transcripts

by the end of this video you will have

the blueprint to become a data scientist

in 2024 videos on this topic are usually

optimized to be as digestable as

possible for YouTube but a 12 minute 25

second video will only allow you to walk

away with a few bullet points and a lot

of confusion today I will be giving you

so much more than that the ultimate road

map I'll not only tell you the most

effective cell taught route to take but

also give you different projects you can

do and in what order at different stages

of your journey the resources that I use

to learn data and most importantly a

bunch of different ways to stand out in

a sea of data scientists strap yourself

in thousands of people get into this

journey they spend countless hours

hunched in front of their laptop away

from friends and family learning all

these different programming languages

just to get to the end and say yeah I

studi data science but I realize it's

not really for me that's what happens

when you don't answer these three simple

questions before you start studying data

science do you know what data science is

do you know the skill set you need to

become a data scientist and does it

sound like something that you want to do

so how do you get the knowledge to

answer these questions you need to do

three different things research what is

data science online just an hour or two

is enough to give you the gist look up a

couple of industries that you're

interested in and see how they use data

science and the third thing you need to

do is look at what a typical day in the

life of a data scientist looks like

research the typical work days are you

happy with the amount of coding they

have to do the amount of stakeholder

interaction they have and the amount of

presenting they have to do for example

give yourself the solid foundation

before diving head first step two dive

head first the two most important skills

we need as data scientists are maths and

programming so how do we pick which to

learn first to me the answer is simple

programming definitely programming

that's because even if you're not great

at maths you at least have a level of

familiarity with it from your high

school days but programming can feel

like entering a whole new world

programming is also how we'll often be

implementing the math that we do you

learn so it makes sense to have this

solid foundation of programming before

you begin the major programming

languages we have as data scientists are

R and Python and don't waste time on

this I'll make it super simple for you

python just pick python your end goal

here is to be as Employer as possible so

just pick python why well when I look at

job postings it's like this 60% ask for

python explicitly 30% ask for r or

Python and very few ask distinctly just

for r with no alternative for another

language so if you choose R you will be

excellently positioned for these jobs

because there is a lot less competition

but the problem is those jobs are much

rarer and you'll be hamstringing

yourself for all of these jobs so just

pick python please the next step is to

pick a simple commonly used IDE a simple

analogy to understand what IDE are if

tomorrow I decided I wanted to write a

book in English I'll have to decide what

software I would write that book in

msword Google docs scrier the software

is the equivalent of an IDE so what IDE

should you write your python code inside

of again I'll make your life easy pick

something people use or at least what I

use vs code congratulations you now have

your basics in order now what do we have

to learn in Python good news I'm not

just going to give you the basics in

Python but also great mini projects that

you can Implement to solidify your

knowledge these six Basics are things

that you use every day and some of them

every single line of your coding Journey

without a good grasp of these it is

effectively impossible to code for data

science so the first thing I want you to

learn is data types and what you can and

can't do with each one of these within

python you probably intuitively know a

few of these such as ins are basically

whole numbers floats are decimals and

strings are just like words and words

and numbers and a combination of

everything in between the second basic

after that that I need you to get a hold

of is a assigning variables and again

this is pretty straightforward assigning

a variable is basically giving your

object a code name that you will refer

to it as throughout the rest of your

code after that learn about lists next

and lists really are super simple to

understand they are effectively ways of

storing different items together within

python so let's say instead of always

referring to items individually you can

use a collective name a simple easy to

understand use case for this is if you

had four variables which were a country

name England Wales Scotland and Northern

Ireland instead of always writing them

out one by one you could just put them

into a list called the UK okay next up

after that is dictionaries and how they

work they really are sort of like lists

except stored in pairs so if you wanted

to include more information you could do

that so instead of just storing the

countries in the UK you could also store

each country and its capital city okay

I'm going to stop giving examples

because it might start to get confusing

but trust me it's super straightforward

it might just just be because this is

your first exposure to these Concepts

okay now we're on to the basics of

Panda's data frames if we want to be

over simplified a panda's data frame is

essentially a fancy way of saying a

table okay a data frame is a glorified

table but you understand how tables work

in Excel okay good well you're at least

on your way to understanding how data

frames work in Python okay are you

feeling confused at this point if so

it's absolutely fine these might be new

Concepts to you and I promise you as

soon as you start learning and getting

the hang of these Concepts You' be like

I remember when I used to struggle with

what a list was trust me progress is

inevitable okay now I need you to learn

just three more things and we'll be done

with the basics and you'll see just how

much we can do with just the basics when

we get to the projects okay the fourth

basic I need you to learn are basic

control flows these are IFL statements

for Loops while loops and all of these

are basically what they sound like it

might take a little bit of grasping but

you'll be fine after this I want you to

learn a basic visualization Library such

as matplot lib Seaborn but personally I

prefer plotly it is just a little bit

nicer looking and it just lets you make

simple charts to visualize your data as

you go along and the last last last

absolute basic that I want you to learn

are functions and how to define and

create functions functions are basically

predefined bits of code that you can

call at any time to avoid writing the

same code again and again so let's say

for whatever reason in your code you're

going to need to divide numbers by two

then multiply by five then subtract

three you could do that manually every

single time which is super time

consuming not to mention boring or you

could write a function where you say hey

every time I give you a number I want

you to divide it by two multiply by

three and then subtract four and then

all you have to do is call that function

and it does it for you much simpler okay

okay I promised I'm going to stop giving

examples because it could get a little

bit confusing but anyway when When You

Reach This stage give yourself a pad on

the back because you've come through

some of the toughest parts of learning

how to code but all of this knowledge

that you've just gained is absolutely

useless unless you apply this one

principle from now on every time you

pick up two to three skills I want you

to implement the principles over here

and that principle is project-based

Learning Without applying these skills

to a project that represents real life

you will instantly forget exactly what

you've learned and you'll think why

would anybody care about lists or

dictionaries that mindset will change

when you apply to a project so these are

three beginner friendly projects to

solidify your skills that you're already

beginning to build the first one is to

create a simple contact book application

in other words within vs code I want you

to create functions that will allow you

to create a contact add that contact to

a contact list find the details of a

contact update a contact's details and

then also delete a contact if you need

to do so the only skills that you'll

need are the ones that we've learned so

far the second project actually Builds

on what we've just learned but it's an

inventory management system so I want

you to be able to create an item with

the price find the details of that item

but I also want you to have a till

balance that updates every time somebody

makes a sale or a return and the last

project is the simplest of the three I

want you to write a function that takes

in an Excel file converts it to a data

frame and then Returns the basic

descriptive statistics of that file and

all of these projects we can do with

just the previously discussed skills now

you know what to learn how to apply it

but where do you actually learn it if

you are going self-taught there are two

options each with its own pros and cons

the first one might actually be the true

self Tor route which would be looking

each one of these up on YouTube and

looking to learn that way it's super

cheap and On Demand but there's a lot of

drawbacks it lacks structure and when

you learn A New Concept from one Creator

they don't know what concept you already

knew before that so they might refer to

knowledge that you do not have yet

that's why honestly I think it's worth

just Shing out the money to get a course

or a a boot camp especially considering

how much you will make once you become a

data scientist the major advantages of

this for me is the structure it will

give you a lot of structure and

importantly you can still get all the

information that you would have gotten

with this method over here and in that

way you can supplement your knowledge

from the course it also has

disadvantages it's not free but still

let's be honest way cheaper than a

degree and trust me I would know and

it's also gamified that was a big

problem that I had with my course and

what I used when I was learning was Data

Camp as it has skill tracks in Python

SQL and anything else that you can think

of I do have my gripes with data camp

but I still use it to this day so I will

leave a link for it in the description

for you to be able to check out the

courses that they do offer and I'll be

making a video in the next few weeks

teaching you how to effectively learn

using an online course so subscribe so

that you don't miss that so everything

that I'm saying might seem like a lot

but if you subscribe to my newsletter

you'll get a written road map of

everything that I have shown to you but

more importantly than just a written

road map is an insights that I will be

sending you one to two times a month

insights that you can only get from

working in the industry and the things

that aren't great for the YouTube

algorithm but will help you to not just

land a data science job but improve as a

data scientist head to datan nash. co.uk

pop your email into the box and you'll

get a road map and a free subscription

to my

newsletter okay so now we have our base

in programming we can introduce some

concurrency into our learning that means

learning two things side by side we'll

be doing more complex work in Python

that will be tailored to making you more

employable and I'll touch on that in

depth a little bit later because the

other thing that we're learning for now

is mats now I don't want you to download

the entire contents of a math book into

your mind so we want the best bang for

the book the fundamental mathematical

Concepts that are asked for by all jobs

so that math includes a basic

understanding of some statistical

Concepts like median mean the mode data

standardization variance and standard

deviation ketosis skewness correlation

and covariant and you can read the rest

on the screen after that we also have

these important linear algebra topics

and linear algebra provides the

framework for many data science

operations and algorithms but the key

Concepts you should focus on are systems

of linear equations vectors matrices

igen values and igen vectors

normalization and distance calculations

after that is some B basic calculus and

trigonometry and you can see the four

major areas I want you to focus on

differentiation integration limits and

trigonometric functions and finally this

is an important one we need to

understand probability and the concepts

in particular I want you to familiarize

yourself with are hypothesis testing

Invasion probability conditional

probability probability distribution and

expected

values take a breath it's just maths

it's going to be okay and unlike High

School where you had to write out pages

and pages of maths by hand and get a

nice thick red X every time you were

wrong the goal here is mainly to

understand the underlying mathematical

Concepts so that you can interpret it

and use it to Aid your decision making

most of the time there will be python

libraries that can do the implementation

on your behalf and your job will mainly

be structuring the code around it and

interpreting those results and remember

you don't have to be a master of these

Concepts but just have a good grasp of

the fundamentals you will be fine one

step at a time remember so what can you

use to help you learn the math well here

are three excellent channels that I

recommend in this aspect stack Quest

redlick mats and three blue one brown

and let's not forget my favorite

resource which is random Google articles

those are pretty good to teach you some

maths now don't forget our number one

principle Project based learning so now

we will combine our math skills and our

programming skills to solidify our

knowledge so the first project can you

code a function that will calculate the

moving average of a series of numbers

and plot the output in a graph in Matt

plot Li after that can you code a basic

statistical function that takes in a

list of numbers calculates the mean

median and mode variance and standard

deviation and for variance and standard

deviation I want you to implement that

manually just to make sure that you have

a good understanding of those Concepts

the third thing code a function that

calculates the dot product of two

vectors for this one again I don't want

you to use any like Li iies at all no

numpy okay for the next one can you code

a function that takes in two matrices

firstly checks if multiplication of

those matrices is possible then if it is

possible multiplies them otherwise it

returns an informative error and then

besides that I want you to explore what

you can do with these two libraries in

particular numpy and

scipi and of course let's not forget my

favorite one the fourth one random

Google articles which have saved me on

more than one occasion we have now

reached the whole goal of our data

science journey and this is one of my

more controversial takes but armed with

just the basics of maths and programming

you should start applying for jobs but

maybe not in the way that you expect I

want you to put together a CV that shows

off your data skills and the projects

that we've done so far but here's the

key part we aren't spending 4 hours a

day applying for jobs at this stage

instead we're just gently prodding

around the market and mainly apply for

entry level roles and internships that

do not require full customization of the

application because right now we

probably won't get the job but Nash

what's the point of applying if we won't

get the job good question two simple

reasons and the second one might be more

important than the first the academic

year of my masters was a 10-month

process but I secured my first

internship to work as a data scientist

in January of 2022 4 months into my

masters when learning data science we

often view the process like this I'll s

of having no skill as a data scientist

and nobody will want to hire me and then

eventually I'll get through all the

courses and get that one final skill and

all of a sudden people will be dying to

hire me the reality is that the skill

acquisition chart looks more like this

where as you gain skills your level of

employability Rises you do not know

where on this axis your first employer

will be willing to take you on as a data

scientist so by podding around with your

CV you get to get your first job as

early in the process as possible the

second reason is all about downloading

data the last thing I want for you is to

spend months picking up all of these

skills and then applying for hundreds of

jobs and never hearing anything back and

trust me this will happen if your CV is

awful by doing a bunch of simple

applications as you learn you get to

test your CV you might put in 20 to 30

easy applications and hear nothing back

and that is a sign to tweak your CV and

see if your response rate improves maybe

your education sector needs to go below

your work experience or change the

length of your CV consistent easy

applications will allow you to to

experiment until you have reached your

optimal

CV with the basics of pipe and mastered

we can now have a lot more fun in this

part of your journey I want you to have

a lot more autonomy around which areas

you dig into firstly by having a

curiosity mindset in addition to this

curiosity mindset I also want you to

adopt just in time learning so as

opposed to just thinking hm I don't know

anything about web scraping let me just

randomly learn how to web scrape the

more effective way is to always be

working on projects and then when there

is a project that demands for you to

learn how to scrape you put aside time

to then learn it to move your project

forward as opposed to learning things

just in case but before digging into the

potential areas of specialization I will

make mention of these three areas that I

advise you have a solid understanding of

well four areas the first is basic data

pre-processing and feature engineering

but I also want you to know what

supervised learning is and the basics of

how to do that the same with

semi-supervised learning and supervised

learn after this common areas in which

to specialize or get deeper knowledge in

the first is natural language processing

anomaly detection predictive modeling

recommendation algorithms marketing mix

modeling computer vision and general

machine learning are all good areas in

which to get deeper knowledge this

doesn't mean you have to only specialize

in one of these at this stage but these

are the areas that you commonly see job

postings for now with all of that

knowledge you should be in a good

position to actually get a job and the

key is to have excellent projects that

that appeal to employers regardless of

which specialization area you're looking

for it can be difficult to think of new

projects that appeal to employers and a

platform that I actually recently

discovered that allows me to think of

great projects is called project Pro

they literally have industry-leading

standard projects and in my opinion are

more advanced than what you can find

online in general so whenever I need a

new project that's the platform I go to

and I will put a link in the description

for it there are a lot of advantages as

you can see listed but the one thing is

that there's subscription price does

reflect the standard of the projects

that it does contain so only do this if

you have the funds too afforded and one

those really really top projects to

stand out quick interjection people I've

actually spoken to the people at project

Pro they've agreed to give you a 5%

discount if you use the link below just

to make it a little bit more affordable

so I think first do free projects the

ones mentioned in this road map see what

you can do on your own time after that

then when you really want to take your

project to the next level if that's

what's holding you back project Pro is a

great place I'm using it mainly for

learning how to implement lrm Solutions

but they have so much more than that so

yeah link below 5% if that sounds

interesting to you if you can't afford

that there's plenty of cheaper and free

options the first one is taking on the

projects that are within your course

this also has a lot of advantages but

also a lot of disadvantages such as

being quite generic focused on skill

display rather than being employable and

often times they spoon feed you to get

through that project but from the free

options what I would recommend is using

your own internal knowledge and

curiosity for example maybe you have a

background in customer service and

decide I think it would be useful to

write some code that would tell you the

sentiment of customer reviews about our

product frequent questions that come up

from customers and the frequently

mentioned reasons for poor reviews now

imagine that you are an employer who

does e-commerce and you see that project

from this person your mind will

instantly think oh wow they could bring

so much to my company if they can

translate that to us because we want to

know what our customers are thinking I

have a whole video here explaining how

to do effective projects to actually get

employed so you can open that in a new

tab or add it to your watch later and to

do any project you will need data so

familiarize yourself with kaggle.com

which is a website where you can get

free data to do your personal projects

with okay so now we have good python

good projects and good fundamentals but

so do a lot of data scientists we now

want to be hyper valuable and pick up up

additional skills that will put us head

and shoulders above the competition the

first of which is SQL which is a great

querying and database creating language

it's excellent and is mainly used by

data engineers and data analysts but

it's still very valuable for us as well

compared to learning the basics of

python learning the basis of SQL is

pretty straightforward but a few key

areas I want you to focus on are how to

query how to create a database including

reducing to 2 NF and 3 NF format working

with relational tables and foreign Keys

as well as elements like creating

temporary tables and some easy window

functions and partitioning again what I

used to learn all of this when I was

going through my soft tour phase was

Data camp and their SQL Developer track

which was actually really good and once

again we've picked up a new skill so

what do we do Project based learning

employers don't care about you telling

them you can do SQL they want to see

that you can do SQL so for these

projects we can have a dedicated SQL

project which is just you showing how to

create a database nothing wrong with it

perfectly fine but option b I think is

integrating it with your existing data

science projects so before we're

building a predictive model on a Cagle

data set now firstly create a database

for that data set reduce it to 3 NF then

do the necessary joins to get the

columns you need to build your

prediction model on top of that I'm

linking this free kaggle data set down

below so that you can do that if you

wish to now the next secret weapon as

data scientists we often do not pay

enough attention to the front end the

customer facing aspect of of our

projects we just concentrate on getting

good at the coding and then leave it to

the data analyst to make it look pretty

but a lot of companies can't afford to

have a dedicated data analyst so they're

looking for a data scientist with the

ability to present their findings and

not just throw a random Jupiter notebook

at them so the next thing that you

should do is become competent with the

visualization software and I do

recommend Tableau when presenting your

work to employers and recruiters you now

be able to show it off both as code but

also as a really appealing dashboard

that Crystal izes the work that you've

done the best part is learning the

basics of Tableau won't take you long at

all so there's no reason not to take a

weekend or two just to learn the basics

now with Tableau Python and SQL in Your

Arsenal and continued work on all three

of these you should be well positioned

to get your first job where before we

were being casual in a job search we are

now being really intentional with our

job hunt really take the time to fix

your CV now and have dedicated time to

apply for entry roles that you think you

can get it's not just easy apply anymore

take the time to customize your CED

where possible for different jobs list

your experience and projects in a nicely

ordered Manner and I will be doing a

completely separate video on this but in

the meantime here's some information on

how to increase your odds of getting

that first job as well as a couple more

videos that I've done around this area

that I will be linking down below as

well listen you will feel stuck at

different points during your data

science journey and if you do go down

this path solo it will get extremely

lonely extreme quickly so you need to

find a community of other people who are

getting into data science or this area

in general for moral support but also to

discuss problems and look to solve these

together you can look for communities

online in the shape of forums provided

by the courses you pay for social media

groups and those sort of things and I'll

be honest I don't have experience with

either of those but I do have experience

with networking on person which is an

amazing resource that I've had great

results with and I do have a video that

exclusively discusses how to network

effectively but the one thing that does

do is limit your community to those who

are local to you and that's a huge

missed opportunity which is why I'm

looking to form a community to solve

these problems that you can sign up for

in this community I'll be having study

sessions and regular calls with the

members to provide more tailored advice

and mentorship on accelerating your data

science Journey it will be for dedicated

fellow Learners and experienced

professionals who don't just want to be

mediocre data sors but want to work

their way up to being truly great it

will provide an ecosystem for growth

support knowledge sharing and it's it's

just a space where you can ask questions

share insights collaborate on projects

and get feedback all of which are

essential to accelerating your progress

if that sounds like something you want

to be a part of sign up for the weight

list below to get Early Access when this

community does go

live if we're thinking about the 8020

rule we're now definitely in the realm

of nice to have things rather than

Necessities but these definitely would

make you one of the outstanding

candidates on top of everything else

that we've already already learned is

learning how to work with apis and use

them to fetch data that can change

dynamically instead of the static csvs

we've been working with when downloading

data off of kago also learn the basics

of GitHub and these Basics are getting

your projects into your GitHub

repository so that other people have

access to your code as well and

something else that I'm learning is

streamlit which allows you to easily

turn your code into an interactive web

application that other people can use

and more on that coming on the channel

soon but that's that's a super useful

skill and the last one this is very much

an extra he is posting about your

journey onto platforms like LinkedIn and

Twitter which are particularly useful

cuz they're more professional at times

and if you have a digital footprint it

shows that you're slowly leveling up and

it could help you to stand out but don't

dedicate too much time to the

documentation at this

stage and now we are on to the final

element which is The Cutting Edge data

science is a field that is always in

flocks so you need to remain up to date

with the latest Trend and the three best

ways that I look to do this firstly

medium and towards data science these

platforms are Treasure troves of

Articles and tutorials insights and so

much more by data science professionals

and enthusiasts whether you're looking

for in-depth tutorials case studies or

thought-provoking discussions on the

latest AI or machine learning techniques

these are pretty good although you do

have to pay a couple bucks a month a

free alternative to this is YouTube

which of course I'm quite biased because

I am on YouTube I think there are a lot

of smart data scientists on this

platform who can give you so much

information so subscribing to a few

channels is always a good idea and the

last thing is following experienced data

science leaders on other platforms again

mainly Twitter and Linkedin those are

excellent resources in order to keep you

on The Cutting Edge and there you have

it the best freaking road map on this

platform and yes this year I'm talking

my so don't forget to subscribe to

the newsletter to get written resour

ources to everything that I've talked

about and at this stage you might be

feeling a little bit intimidated

wondering if you have wanted taste to

become a data scientist I have this

video over here that addresses whether

you are too dumb to be a data scientist

so click on screen now