Gigs: A day in the life of a data scientist

[@davidbasil2727]

- What does it take to be a data scientist? - "A lot of imagination".

[@00fabian7]

Take the 'heavy math background needed' with a huge grain of salt. Of course you need an understanding of how the algorithms you apply on your data work, but the hard math aspect of it gets done by the functions a package brings. You will need some affinity with maths, but not as much as this person would like to tell us (possible because of his engineering background). Also, here was said that they mainly only use Python/R, but if you're going to work with larger data sets (which is something that gets more and more attention in the future) you will also use programming languages like: Hadoop (MapReduce), Spark, NoSQL etc. As last point, data science has multiple aspects that you can focus on, for example: legal business (regarding law, privacy of data etc.), business infrastructure/logistics, optimization/forecasting, visualization/communication and so on. Somewhat more ignorant employers will want one data scientist doing all of the above and more ('unicorns') but the larger companies will oftentimes have you work in data scientist teams of up to 10 people.

[@GodsNode]

Yeah the math and stats knowledge will be on a conceptual level and the knowing all the inputs and how they fit into the equations.

[@ramiahmad1973]

Fabian

[@rerebacchus]

just as a clarification -- hadoop, spark, and nosql are not programming languages!

[@sessarichard]

Not really. Any real data scientist will have a solid background in the differential/integral calculus, mathematical statistics, linear algebra, and more. People think they can just read a book and learn data science in 6 or 12 months. Doesn't work that way at all. True data scientists usually have a PhD or postdoc as a prerequisite .

[@00fabian7]

True to a certain degree. I do see now I've phrased my comment quite bad (equaling hadoop etc. to programming languages, brain aneurysm I guess) I'll just hide behind English being a secondary language of mine. Anyhow, it's true Data Science isn't just a 'let's watch some youtube videos and I'll know all'. The minimum for a decent Data Scientist is in my knowledge a MSc, with indeed many also getting a PhD but I've yet to see hard requirements for it. It's mostly asked what capabilities you have (knowing about statistics, machine learning, data visualiation etc.) next to programming requirements (python/r, hadoop etc.). And maybe because of my bias as being next to BSc/MSc of Mathmatics whom I then compare to regarding the degree of understanding of mathmatical concepts. You don't need That level of maths (although that also depends on what role in a data science team you want to uphold). In short, a PhD is not the prerequisite to being a 'true' data scientist, but a thorough mathmatical understanding is needed (though not on a level similar to a BSc/MSc in Mathmatics). Personally I see a MSc to be a minimum to be decent enough as a starting point. Or an equivalance in experience and skill (with the personal qualities to be able to convey that during an interview without a degree to tell that for you).

[@HarshReddyGandavarapu]

He took a U-turn in career. Phd in solid state physics and now working on Data Science. That's interesting.

[@luvsuneja]

Not really a U-turn, still tech related. I've heard some Engineering PhDs become real estate brokers.

[@mpete0273]

This is common. In New York City all the data scientists are physics and math PhD's.

[@hottguy1005]

Hey guys I want to get into data science I already have bachelor's of science in computer science what do you guys recommend me to do masters in? Thanks

[@KurtGodel432]

Hardly a U-turn when he’s now making more money than he was as an engineer.

[@DrLivingParadox]

Don’t forget he also had mechanical, electrical, and industrial degrees/backgrounds too!

[@Akhil2481]

thanks for showing me all the email addresses at 2:30...that's data breach at ATT...did no one notice while editing the video?

[@bharatbshetty]

A M 4th of February 2015...lol

[@Nedwin]

lol

[@sargebeats]

yo what song is that!? that guitarist is shreddin' it! \m/

[@dun-kindohnatson6447]

Very interesting video. All of your other videos are great too! They show great insight on the many interesting, rare, and well paying careers out there!

[@ArifinRahman]

Nice interview, answers a lot of rookie questions. Thanks!

[@harishaw9988]

Hey I have maths background as I completed my masters in Maths but not sure how to start with Data Science. Can anyone please suggest??

[@MrMiles333]

I am doing a website for teens and what kind of technology careers are out there and this video was great and the music was perfect for what I was looking for. it will keep the students interested and not bored! thanks

[@josephjoey3904]

What does he mean by Maths background? Can someone kindly elaborate on this, thanks

[@sommoimtiaj7138]

is it difficult to become a data scientist if anyone not from the math/engineering background?

[@JoeG2324]

Pure nonsense regarding math skills for data science. I work as an sql developer and work heavily with sql and I work with data scientists who spend most of their time working with tools like python where all the hard is already done for you. Yes, you do need to understand the models and interpret the results, but this guy is a gigantic bullshitter in saying you need a math background. You need to know stats, but you don't need to be a PHD in stats to be good at data science.

[@Tntpker]

Joe JG Well I think that he meant in a way that if someone that really despises math and is thinking of pursueing this carreer, should better not even try.

[@js913]

Joe JG and google

[@lokikingofasgard5411]

Import pandas, numpy,sklearn That's is just python and idea wat u have to do with data

[@kartik3088]

I disagree with you. I use R on regular bases to solve problems at work and it is essential to have a solid math foundation. For instance even if you were to use a basic algorithm like a linear regression you’d need math to understand how the back end of it works I.e How your algorithm is calculating residual square errors which intern helps you understand its accuracy. Yes, the software spits all those values for you but if you want to explain leadership what those values mean. How would you do it if you don’t know the math behind it?

[@lokikingofasgard5411]

kartik mutya yes I agree with u but we don't need PhD for data science jobs ... 3 months training is enough . No one ask u to implement u r won ml models....90% of time we use existed ml algorithm models ...if orgs wants new ml algorithm model.then we need bunch of PhD guys

[@pratthameshchakorr2834]

I am aslo a mechanical engineer. wanted to switch to data science. is it ok?

[@Jacob.Cornejo]

Awesome music!

[@jmatt98]

is this a new trend cat6 hanging from the ceiling?

[@relly793]

lmao

[@manicsurfing]

its feng shui

[@AhmedAbdulSalam5]

If u see this comment.. pls do reply.. Have u tried youtubing the carrier path for data scientist..? Aka.... things u need to know.. also.. math heavy background is not that heavy.. all it takes is u to know how u can logically think about numbers and interpret them in the most logically correct way.. just like ur flight seat number.. when u see it.. u already know where it is cauz u correctly interpret it.. same is here...u have tons of numbers... U will know where they can from and also what they mean.. now what u have to do is to make something interesting out of it..like.. generating a best possible solution based on clients problem.. Aka.. u see the flight booking system..or let's say seat reservation in bus.. that use numbers and alphabets to locate the seat right..? Here is the thing.. client comes to u with all those numbers and asks to book all the windows seats of a bus with one click.. so what u do is get the info from him about what those number mean and get the good details on his problem like.. what kina bus.. dates of the busses like soo..and more info like taxes etc.. just as much as required.. now what u do is to make a simple algorithm that will book all windows seats.. and give it to client..what did u do here..? U played with numbers cauz u r math heavy background u know!!! This is how it works.. random numbers... That relate to something in real world.. u use some algorithms to make a solution and give it to client or push into a robot or a PC to automate it.. more like a software that can detect and book all windowseats of any given bus...voila... This is just a simple example of how things work here.. bt in reality..trust me there will be TB's size of numbers and u handle them with and gentle look on them.. most of time.. u r using a interface that will automate work for u.. using that. U make a automated solution... Yatatatatatatatatata Common. Man I close my csgo to learn this things.. now just think how interesting this field can be when u create a bot that can think and shoot instead of shittty aim hacks and wall hacks shit For more motivated example.. Try googling "python plays gta5" Rock it babz!!

[@positivemindset8710]

What are the other skills apart from these to be learnt to get into this job?

[@shaileshtiwari8289]

I have a bachelor's degree in finance can I go for data scientist..?

[@WillyJunior]

2:32 - did they really just publish all those peoples email addresses?

[@dheerajbh]

Can a arts student pursue Data science?

[@ibrahimsani9718]

great interview on data science,i dont know why people are missing the whole point of the video and are focusing on the music.ughh for the sake of sanity this is not a music video.

[@beamjohn9753]

Good video! Hope to see more like it in the future on this channel.

[@mrswolls]

Why was this edited by the high school av group in 1996

[@afaqansari149]

Interesting talk, crappy music. Almost gave me a heart attack.

[@mooghero909]

Afaq Ansari what are you, made out of glass? Lower the volume if it's too loud.

[@nightfire4107]

😂

[@sudeepkadyan5480]

@@mooghero909 lol

[@reneewoodruff5972]

People are getting riled up on the whole 'you need a PhD to become a data scientist' thing. You don't need to be a PhD. What this man infers is that having a PhD in a math heavy background (engineering, math, physics, sciences, CS, etc.) will make you a great data scientist. Data science is all math and logic. Sure, most data scientists hold PhDs but a lot also hold master's and some even just bachelor's. If you want to become a data scientist, you shouldn't focus on attaining the highest level of education possible, but make sure you integrate your studies with lots of math and stats. There are even self-taught mathematicians who work as data scientists.

[@cristhianmalakian296]

The music is ok, though old people prefer calmer music

[@praveenm1001]

Maths and computer science background are an added value for sure. But not necessary, you need to understand basic statistics, have an inclination analytical thinking. Thats it, rest all you can learn on the go be it R python anything. Yes experience in any of these will jumpstart you. This video definitely shows the difference between the way a person who "does" the stuff talks vs a person who just sits on top reading presentations :P

[@BEPEC]

well explanation , loved it

[@tipusultanbargir5888]

Sir I'm civil engineer ... Can I change my carrier into data science... & Which cauurse can I do

[@vinodvarma3405]

Great learning, upx academy these academies will provide online classes search it on google make a contact with them👍

[@arsalaanayoobi8349]

do this explains that if we are data science as the video

[@johnyvicewax7375]

Thanks Joey Jackson

[@cwaddle]

I think that the job of a Data scientist is very much like that of a statistician. I guess that with a data scientist, you are sort of expected to do more data manipulations and coding, creating an engineered solution that works with a heterogeneous large data set

[@iLoveTurtlesHaha]

Oh man, I want to get into this field but I suck at Math :(

[@johnnycincocero]

iLoveTurtlesHaha You may think that is the case, however, you're just not working hard enough!

[@AhmedAbdulSalam5]

If u see this comment.. pls do reply.. Have u tried you tubing the carrier path for data scientist..? Aka.... things u need to know.. also.. math heavy background is not that heavy.. all it takes is u hlthay how u can logically think about numbers and interpret them in the most logically correct way.. just like ur flight seat number.. when u see it.. u already know where it is cauz u correctly interpret it.. same is here...u have tons of numbers... U will know where they can from and also what they mean.. now what u have to do is to make something interesting out of it..like.. generating a best possible solution based on clients problem.. Aka.. u see the flight booking system..or let's say seat reservation in bus.. that use numbers and alphabets to locate the seat right..? Here is the thing.. client comes to u with all those numbers and asks to book all the windows seats of a bus with one click.. so what u do is get the info from him about what those number mean and get the good details on his problem like.. what kina bus.. dates of the busses like soo..and more info like taxes etc.. just as much as required.. now what u do is to make a simple algorithm that will book all windows seats.. and give it to client..what did u do here..? U played with numbers cauz u r math heavy background u know!!! This is how it works.. random numbers... That relate to something in real world.. u use some algorithms to make a solution and give it to client or push into a robot or a PC to automate it.. more like a software that can detect and book all windowseats of any given bus...viola... This is just a simple example of how things work here.. bt in reality..trust me there will be TB's size of numbers and u handle them with and gentle look on them.. most of time.. u r using a interface that will automate work for u.. using that. U make a automated solution... Yatatatatatatatatata Common. Man I close my csgo to learn this things.. now just think how interesting this field can be when u create a bot that can think and shoot instead of shittty aim hacks and wall hacks shit For more motivated example.. Try googling "python plays gta5" Rock it babz!!

[@MrGarg10may]

Just a talker. Probably a manager.

[@prighwaitnaamee2193]

Tanmay Garg lol yeah

[@nishisaxena4831]

Tanmay Garg yeah he sounds fake ...

[@rakiraj87]

He repeated heavy math background three times as if he got nothing to say

[@maddy95kk]

thanks for the video. it helped a lot as I am trying to find a job in data sciences

[@kshitijkulkarni8698]

Thanks

[@TheTraffic247]

with all the degrees doesn't sound like he knows much. seemed confused. you barely need his background to become a data scientist.

[@johnnycincocero]

How do I start? Are there degree programs tailored to becoming a data scientist?

[@043mehdi]

to be a proper one. Statistics. Business analytics. Business Intelligence. YES THERE IS. Do you need more

[@johnnycincocero]

@@043mehdi YES

[@1anre]

The music is trash. Please don't torture us by using such music again when presenting intelligent subjects like this.

[@RashedHiyasat]

i completely agree, i can't continue the video because of it

[@akshaybachhav8768]

absolutely right sir

[@KillahBeeSwarm]

but its RAWCKING yeah! alright! \m/

[@asiddiqi123]

We need a *Rock Star* Data Scientist

[@relly793]

lmao

[@blaby4ever]

You could of put the spongebob music, everyone would of been okay with it

[@nitishsonkar3269]

data need more background of statistics and other coureses.....give right advice atleast

[@peaceful4you408]

You don't need a heavy math background for data science. I don't know what this guy is talking about. The only math you need is statistics and that's if your doing the math aspect of data science. Data scientist can get involved in coding, math, and many more other things. I worked as a data scientist that did nothing but coded in SQL before I decide idea to become a full fledged SQL developer. Now, I deal with nothing but database work and nothing else. Data scientist can do anything. It depends on the company. If you want to be a data scientist that just deals with programming I would suggest learning SQL and Python. That's it. If you going to be a data scientist that wants to be involved with math, I would suggest Statistics and Calculus. Data scientist work in groups alot so by knowing different things can bring better productivity to a group instead of all the data scientist having just a math background.

[@iLoveTurtlesHaha]

I don't know man, I feel like I'll suck if I didn't have some Calculus background in there. I'm more interested in creating algorithms for deep learning but I guess if you go that far, it's no longer data science? Not sure. It's kinda confusing for someone just starting out. :(

[@peaceful4you408]

No. You'll just me more of a marketable data scientist I'm sure. The more technical skills you have, the more desired you'll be. Companies love data scientist that is competent more with engineering and math than say programming or machine learning which are my skills. I have been able to get by with just these skills so far. It depends on the company. I was with Apex systems at the time. I'm now at a startup company and they treat me as a genius here. Those who are technically inclined know my skills are mediocre but to those who knows nothing about technology, they think I'm a genius. This perspective has had an influence in my salary and level of respect at my current startup company in Washington DC. I love the role Im in now. Just a little worried about job security because I don't know if this company is experienced enough to make it.

[@dcjohnson2490]

I can't tell you how relieved this post makes me feel. I'm not good at math at all but definitely have an interest in data and making it a career. I have knowledge of SQL and I know Python is almost a requirement for data analysis. Do you think I should try to learn both languages concurrently? Or continue on with the SQL and then move on to Python afterward?

[@JordanShackelford]

Statistics, linear algebra, calculus, programming knowledge

[@timcook3410]

Ryan Heard what is the purpose of a data scientist?

[@adebayooyeleye6410]

Nice video. Thanks!

[@deeplearningpartnership]

Very interesting.

[@gauravsharma-mi2er]

Music perfectly matches to what he is saying.Good Job !

[@bigspook999]

What country is this

[@saitaro]

Somalia

[@HappyDataScientist]

Great Interview ! Thanks :)

[@SantoshKumar-ou8jr]

Thanks !!

[@oliverli9630]

At 1:38, there comes a man from the left of the screen. does data science turn him into that? :O

[@yashvardhansingh5681]

Thank you so much for this video its vry hlpful

[@j2klegend]

hmm at&t company mail ids!!! wow

[@satishbanka]

Nice short video!

[@positivemindset8710]

dude, u can learn data science from institutes as well. I enrolled in an institute named Skillathon in Hyderabad. They are teaching fine. Go have a look at it @ ravi shankar gowd

[@ImGooblie]

only someone who knows nothing about data and is a mech engineer would try to define data science as a math problem.

[@043mehdi]

damn son, you r right on point

[@9zQx86LT]

Math Backgroud Math Background Math Background ! :D .. I have a commerce background.. and I am an Aspiring Data Scientist.. Take This ! ;)

[@PrebenOlsen90]

Hi Ashish! I too have a commerce background, and I want to see if a career as a data scientist/analyst is an option for me. Could you lay out some words on your background and how you approached a career into data science?

[@web3dopamine]

Yes. U can. please inbox me

[@mpete0273]

Unless you're coding models from research papers or implementing new algorithms yourself, I can assure you that all the math you need has been done before, so just import it.

[@narenspirit]

Hi Ashish pls share your email id....need some guidance on data science...pls

[@NorthTexasEagle1989]

Plano reppin!

[@linglee5759]

Need a PhD now for this stuff. Same as a quant.

[@vikramchavan1513]

After 1.5 min I realised why I am here obviously a bouncer

[@saurabhtripathi62]

Background music u use is extremely irritating

[@shawnz9833]

cool

[@hypnotist1629]

you are funny and good

[@ben456tone]

I'm sorry Math for what???

[@nightfire4107]

3 degrees in engineering & a PhD in physics!? Dude is smart

[@MrTuushar]

I hate the music!

[@punnu5997]

Your background music is fucking annoying and completely unrelated to the mood of the video. Try and put in relatable and mood matched music.

[@fox971]

This kind of people only pretend as most of them.......

[@js913]

Engineering is my math heavy degree. Lol

[@jeathtunes769]

I smell chicken bhuna

[@akashdeoraj7699]

Jeath Tunes I smell mash potatoes looking at your face

[@magiksity]

damn what a stupid music choice

[@elinelindazagovawintersong708]

Bulshit......

[@nipundudeja]

Saale indian hoke bhar jaak3 kaam krte ho 😤

[@WillyJunior]

I work around the indian hawk.

[@griimrippper]

Traitor

[@QSing999]

Sorry but this guy is taking a little bit of truth and stretching it far more than he should. He is no scientist, he is a bull shit artist. I am an engineer and know engineers and this guy this grand standing.