The TRUTH About Learning Data Science
Vložit
- čas přidán 4. 10. 2021
- Subscribe to RichardOnData here: / @richardondata
In this video, I'm going to talk about all these "learn data science in 2 months / 3 months / 6 months" articles, courses, and videos that you see out there. I want to talk about the reality of learning data science and what it's more realistically going to look like. You want to think of it as a marathon, not a sprint. Learning is not a singular act. It's a mentality, not a process. And that's what will set you up best for the most fulfilling and lucrative experience with data science possible.
Coursera statistics courses:
Duke: www.coursera.org/specializati...
John Hopkins: www.coursera.org/specializati...
University of Amsterdam: www.coursera.org/specializati...
An Introduction to Statistical Learning: amzn.to/3k8SaRN
The Guru's Guide to Transact SQL: amzn.to/2IgvGko
R for data science: amzn.to/3je8kK6
Python data science book: amzn.to/3cDXKcE
Linear algebra: amzn.to/2HEj4U4
Tufte book (data visualization): amzn.to/3kVrR2o
Few book (data visualization): amzn.to/3n2qTTU
The Hundred-Page Machine Learning Book: amzn.to/3563PfK
Applied Predictive Modeling: amzn.to/38lbf0M
Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: amzn.to/3eAh1N2
Approaching (Almost) Any Machine Learning Problem: amzn.to/36gSjh0
Deep Learning: amzn.to/3p83vVW
A Study Pathway for Data Science: • A Study Pathway for Da...
Should You Be a Data Science Generalist or Specialist?: • Should You Be a Data S...
PayPal: richardondata@gmail.com
Patreon: / richardondata
BTC: 3LM5d1vibhp1F7pcxAFX8Ys1DM6XLUoNVL
ETH: 0x3CfC599C4c1040963B644780a0E62d45999bE9D8
LTC: MH8yPjvSmKvpmRRmufofjRB9hnRAFHfx32 - Věda a technologie
You hit a nail in the head. Data science field is full of influencers who are trying to sell an idea of being great at it within 6 months which isn’t a case at all. It isn’t like learning web development where you can learn basics within a short period of time and get a job. Statistics is a complicated field and you really need to have a deep understanding of it to apply it efficiently and make judgements about which methods you should use.
Yep, exactly. I like the idea of having principles and process for learning, but having unrealistic expectations is you end up demoralized down the road!
I very much agree. I'm actually getting tired of seeing so many influencers who may or may not be knowledgeable over simplifying just to get views and likes.
I Have been learning Data Science for 10 months, and it just started to make sense 😂
Welcome back Richard. Keep on Julia tutorials ! 👍
I started to really study after I graduated from grad school. Learning is a lifelong process. Nice video!
Yes! Thank you!
What is your degree?
Annica best of luck!, You would want to study the hardest you possibly can. Algrebas and calculus and Statics and python and sql programming language and maths give ur best it can take 3-4 years so keep on going !
Great video, Richard, keep it up!
The way you talk is very unique and rhythmic
Great video, some influencers really do a disservice to others by emphasizing the destination and not the journey.
Another great video, Richard. One reason I’m interested in this field is that there is SO much to learn and so many opportunities that appeal to me. I’ve actually thought about starting out in data analysis and then perhaps growing into more of a data science role, because I really enjoy learning to code and I loved stats in college. Question: Would your answer change if it were a data analysis path vs. data science where I wanted to begin? Thanks again for your guidance.
Distilled wisdom, thank you
My pleasure!
Great video and information, Richard!
I'm trying to make a career change into data science. I have an IT degree and a master's in astronomy, both of which have been immensely helpful for learning data science. But even with skills in coding, statisitcs, and scientific thinking, putting it all together requires a learning curve as well.
astronomy is cool! I hope you can teach me how I can get to that career path or what do you do as an astronomer
Data Science is a broad term applied to everything data related. Most companies need Data Analysts and Data Engineers. Most need to efficiently and securely identify, collect, transform, analyze, and report on various pieces of data that affects the company. Some may need Machine Learning which can be used to inform strategic decisions based on the data.
A Data Scientist is different. A Data Scientist researches, tests, and experiments with data and develops new algorithms that can be implemented in applications by developers. It's primarily an R&D role, and unless the corp. has a large R&D budget, it's useless to hire one. Most R&D projects fail, but the one that hits can transform an entire industry.
I don't want to know the truth. I want to bury my head in the sand and live in my own fantasy world.
Hah, well self-awareness goes a long way!!
I'm very happy to have gotten data analyst job using r. I think it could be a good catalyst for data science job my sup said I would be doing some ml in the near future. Just graduated but still trying to master caret and the tidyverse.
Great video Richard! I recently had a wake up call when I had an interview for a data science role and the interviewer explained just how intense the data science interview process is for this company. I’m currently in a masters program which I realized is still not enough. There’s a lot of learning/skills I’ll still need to learn before I could really make it through the technical gauntlet that is a data science interview
Yes, absolutely. And don't give up, because some are much more difficult than others. Ultimately, you never want to work a job if you legitimately don't think you can perform it.
Can I ask what the interview process was? Good luck!
@@bulletkip this was a couple weeks back but from what I remember they said it required an online coding test, a technical interview, and a mock presentation in front of some of the department PhDs. They also said they would look at my github/portfolio which I don’t have 😬
The process for data analyst was just two non-technical interviews.
@@ricand5498 thank you
Even though it’s an older video, I’ve found richards comments to be true. I’ve been learning for 21 months and have decided to start a degree in data science. It may take me 5 -7 years to learn data science, in the end, enough to be qualified for good beginner jobs.
I've been studying data science in a Masters since an year and half and I got my first job three months back ago... the gap between courses and real life data science problems is huge, the problems I face in my job are way too hard so the imposter syndrome has been strong these months.
I started just couple of months ago. I'm giving myself 3 years.
3 years should be plenty!
Im a 34 year old guy in the area of finance / audit. I ended up working part of my assignments as a data analyst and my interest and desire to change to a full time DA/DS career is now stronger than ever. What is your opinion about a career change for data science for people from a non-stem background (such as a business management major for example) ? Most of the examples I hear are from people who were in related fields such as Math, Physics etc. Im really considering the feasibility of it before making the jump. Thanks for the content!
Business management could give you an excellent background for data science, because frankly it has many who lean too heavily on the side of being too technical but lack a lot of the necessary business and domain acumen to be as effective as possible. So sure, there's some learning to get there but I think it's completely feasible!
They say that the more you know and learn, the more you are aware of how much there is to know and still to learn. Someone who tells you that you can learn DS in six months simply doesn't know enough to know how much there is to know and learn.
What kind of DScientist do you want to be? I want to be one with a very solid foundation in Stats working on machine language on embedded or IoT devices and work with sound stream data.
I am working on an MS in DS and a grad certificate in stats at the moment.
I've been learning math and python for 1 1/2 years and still can't do data science. Bought a posgesSQL book a week ago and I can't complete it because it won't download the dataset they write I need in order to do the 300 pages left in the book books. Today I'll be spending up to 3 hours just to find a credible Good SQL course. I've done courses (like 365Data) that are NOT hands-on. Thankfully they do have a proper refund process. I know enough to know what's crap and what's not and Richard is right - the idea of learning data science in a year is total BS unless you already know statistics and have learned another language well beforehand. I am expecting it to take me 1 more year. That means it will have taken me 2.5 years to learn the basics of data science. Note that I had kindergarten level math and no previous coding experience when I started.
Which PostgreSQL book did you buy and what is the URL of the dataset?
I like your videos.
can you make a video to students about how could they be qualified for there first job after graduation ?
You know what, this is a great idea and might just be my very next video topic!
@@RichardOnData Thank you so much.
Some of the books you have recommended have a bad review … Can you do some book reviews?
I think the same as you !
licensed career counselor ?? is that a real thing ?
Yes it is, there are certifications and whatnot for such a thing.
Here's a simple example. I routinely see people use clustering and partitioning algorithms. Do they ever check that the data actually has clusters|partitions with a simple Hopkins stat test? No. Do they ever check the p-values of the clusters? No. Or the cophenetic coeff to test the quality of the clusters? No.
I also see people who confuse and conflate linearity with collinearity thinking that it's one and the same. Quite basic stuff that people simply don't know. Just examples of the amateurism out there.