How to Become a Machine Learning Engineer in 2022
Vložit
- čas přidán 26. 07. 2024
- Here is the step-by-step guide on how you can become a Machine Learning (ML) Engineer this year!
If you enjoyed this video, I would be excited to connect on Twitter or LinkedIn.
Twitter: / datawithsandro
LinkedIn: / sandro-luck-b9293a181
Medium: / datawithsandro
0:00 Intro
1:13 Learn The Right Skills
10:05 Where to Learn The Skills
12:28 Create A Project Portfolio
14:49 Create A Resume
17:06 Linkedin is Your Friend
18:00 How Much To Learn Of What - Věda a technologie
Most valuable video of 2022 for me. Appreciate it Sandro, this is amazing stuff and really well explained!
Extremly happy Yalslaus this video helps you as much as I hoped!
Super valuable info, thx!
Super happy, it is valuable to you!
Very informative video! Thank you so much for bring knowledge for us! 🇧🇷
Glad it was helpful!
Its 2023 and its still a great video for me starting AI journey
Thank you, it is a great video because this represent a source of far for bigginners
Very happy you like it!
thank you
great you enjoyed it!
and what about NLP engineer? I’m actually thinking to go for a master degree in computational linguistics and language technologies and I’m actually getting very interested in NLP and help machine’s understand language, I know there are many linguist that work as NLP engineers, and also I even saw many job announcements about ML/NLP engineer that were asking also for a degree in computational linguistics ( others even a PhD in math so I think it really depends ). But I’m not sure if this is a path meant for linguists or not. Do you think NLP is a job that you can do only if you come from a STEM education with solid knowledge of math -algorithms? Obviously I’m not referring to ML/NLP researchers that research and create new algorithms and models from scratch, that’s not even what I want to do actually, but I mean to be an “applied NLP engineer “.
Ppl told me you actually don’t need much math if you don’t want to create new algorithms, you just need some knowledge of calculus and linear algebra ( and more stronger knowledge of statistics) because nowadays you just use PyTorch or similar, but I don’t know until what point it is true
I mean, for sure I know that I can’t take a ML university class in my degree since it requires too much math since it’s the classic theoretical academic course, so I’ll learn ML in other ways
Excellent Excellent Video! Whenever I search this topic, there are always two categories of suggestion: 1) to learn ML extensively including the math (ESL book), implement ML projects, ML papers. 2) the software engineering path + ML.
Are there any differences in the type of jobs? Could you also make a video on how to run ML algorithms on a budget. Should we build our own rig? Is it worth paying 200-300$ for every project?
Good question, I would say in theory the people studying more the mathematical side might lean a bit more to those jobs where a strong statistical foundation is worth a lot( let' say finance, insurance etc.) Those with more software engineering might end up in more technical ml projects like how to make a model that can predict in X seconds( many of the software companies). In practice it will also depend a lot on what you do with your skills on the job, e.g. which industries you go into what projects you work on over time. The job titles will be the same mostly something between Data Scientist and Ml Engineer, maybe Ai Researcher or Ml Ops Engineer( differences are often more from company to company than a general thing. Personally for the budget part I would by now say either if you already have a gaming PC use that one, should be good enough for most projects. If you are working mostly on a laptop, I feel Cloudbased notebooks will be a lot cheaper unless you run those GPUs 10h per day. Let's say GPU+PC=1200$ on GCP you pay for a T4 with a jupyer notebook around 0.5$( give or take) means you can run this for around 2400h (like 100days straight)
@@datawithsandro2919 thank you very much for the quick detailed response, Sandro!
There is one youtuber name datajanitor in his video he says data science is just about data cleansing and some statistics real works are done by ml engineer and he also says that u r not gonna get a job as a mlm engineering if you are a fresher
So to get a job in ml engineering you should have 3 years sql experience
What is your take on that
Well that 3 years Sql part is a heavy overestimate, but yes you need to know Sql and data cleaning is extremly important. Data preparation in the end is where you spend a lot of time
So in general yes learn sql, but don't learn that for 3 years( many things by now are optimized away, that used to be important 10 years ago)
Why I didn’t find this channel earlier.
Lovely to hear!
I appreciate your effort on creating this video Sr. Currently I am a data scientist, do you think Scala is worth learning to become a ML Engineer? I have looked on several posts that this is a must. Thank you.
Thank you so much. In short no, the dominant languages are python and occasionally C++. While I learned scala at some point, I think it is very rarely used and if it is in context with spark/sparkML. This can also easily be done using pyspark. Its useful but surely not a must
just to add, if you are intrested in really big data ML spark can be extremly cool( where i guess also this scala focus comes from)
@@datawithsandro2919 I really appreciate your help. Thank you 👏
Now I want to go to computer science degree in software engineer program, I want to concentrate on machine learning and become a machine learning engineer.
I believe that machine learning and deep learning will soon move into real world, they will not be only on the computer. I mean, there will soon be self-driving cars, robots, drones, and all of them will use neural networks and deep learning.
My question to you:
Is if I go for a degree that is purely related to software development only, can I apply machine learning to the real world in my practice, such as self-driving, robots, computer vision, etc?
What do you think will happen in the future, will machine learning real world applications become more widespread than it is now?
hey there fully with you, and something I personally also looked into in the past. The question seems to be a bit if computer science or robotics is the better field of study for you. Generally both are great and offer both options in the respective other domain. In the end I think you can start in both domains and switch to the respective other. For example I am pretty sure Tesla employs both types, I think you have to decide if you are more into the actual interaction with mechanics(more robotics, mechanical engineering) or for example Vision/Geo Location algorithms( I feel here it is more computer science). It's not a full answer since it really depends on what type of robot/machine/position you are looking for, both domains have to collaborate to truely bring robots to life one day
Second part: Yes I truly believe some form of ML will be part of many applications to come, stretching into the real world with industrial robots, self or semi self driving cars and all that other cool stuff in scify movies( that movement has started and surely here to make some big advancements in the next 30 years)
@@datawithsandro2919 I am absolutely more interested in computer science, I see myself as a programmer, I am not interested in robotics and mechanical parts, I am interested in the software itself and I want to dive deeper into machine learning and concentrate on it.
I'm talking about the real application, because I have in my example a person like Andrey Karpaty, who is essentially a programmer and scientist in the field of AI, and he was working on computer vision at Tesla, which is essentially a real application of AI, this is the field that I most Interesting
This is a great video! I am just starting my bachelor at WGU and venturing into understanding the world of IT/AI. I am not sure which of the degrees offered by WGU best suit the machine learning engineer goal; software engineering or computer science? Do you have thoughts on this? Any additional items I should focus on outside of the bachelor to ensure employability?
One thing I'm very confident to say is, the degree is never enough. I know a lot of people who have bachelor's degrees in CS/ IT and were unemployed for years.
I am doing BSCS at wgu. Its not enough for MLE. We need more education and experience to enter MLE field.
Hi Sandro, is the 360p maximum quality??? or it's a problem from my side??
Sorry for that, I released it before the HD processing was done. Now it should work👌
Informative video , thanks. I am computer science undergraduate university student, interested in machine learning. Do organizations expect master's degree from candidates ? How good I have to be in statistics ?
As often there are many ways to achieve your goal. I think you will be fine with some basic knowledge to get a job in the area, especially if you focus a bit stronger on MLOps at first. While a strong background in statistics is essential for the more data science heavy modelling work, I am sure you can pick that up along the way. I do have one more video on how much math you need that dives a bit more into the specifics, best of luck
While a msc degree used to be required and largely still is( we recently hired and all applicants had a MSc) I think companies are going a bit away from that requirement. As we usually say master or equivalent work experience, that you might be able to pick up at a data driven start up
update it for 2024 then, plz. Worth to become ML engineer in 2024?
I am a frontend developer. Do I need to learn backend development to become a Machine Learning Engineer?
Great that you decided to do it. Now this depends on the specific position, some are more devops heavy and some more backend like api,sql,keyvalue and what not. In general yes it is useful to get good at devops and general database systems( depends now a bit what you understand under backend). But also frontend is often useful when building small applications for internal users
Do you do model building?
I would say yes i do, but it depends a bit on what 'model building' means. Do i design fully new Ml architectures from scratch -> no. Do i build ml systems that use multiple state of the art algorithms and combine/train and put them into production -> yes. For the rest in between i would say sometimes
@@datawithsandro2919 I ask this because in some places, the MLE is also the Data Scientist.
So, in your case, your work with data scientists?
🧐
Is it possible to get into Google with this road map sir 🙏😍
Generally yes. I have recently talked with some friend from Google apperantly while hiring they put now a lot of focus on skills (especially their own Skillbages/exams, checkout Qwikilabs), education(up to BSc. is considered), your skills how good you are in all relevant areas in this video, for google you will probably have to specialize also a bit more in large scale ML. But this is just information I got over a beer so double check that maybe
Sandrooooo… where are you brother? Come out of CZcams retirement! Hope everything is okay