How Much SQL Must I Know for Machine Learning Jobs?
ฝัง
- เผยแพร่เมื่อ 14 มิ.ย. 2022
- Let's talk about SQL, Machine learning and what you need to know.
LogikBot - Affordable, Real-World and Comprehensive - www.logikbot.com
There are two core paths on LogikBot. One for the data analyst role and one for the machine learning engineering role.
I've added the data analyst role recently because machine learning has become so difficult to break into. If you're new to IT, I'd highly recommend this path. You can be in a real-world data analsyt role in under a year if you really apply yourself. LogikBot is the most affordable and fastest way to attain this role. The platform even has an exam simulator for preparing for the exam.
For the data analyst role, you take two courses then prepare for and pass exam PL-300. This exam, PL-300, is the Microsoft Certified PowerBI Data Analyst Certification.
If you're in a shortcut role, then there's an entire path laid out for learning machine learning. The platform used by Uber and Microsoft. I worked at both companies. The courses are real-world and comprehensive. This is not the easy way. There is no easy way. This is one of the few platform that's offers a real-world approach to working in AI.
LogikBot - Affordable, Real-World and Comprehensive - www.logikbot.com
There are two core paths on LogikBot. One for the data analyst role and one for the machine learning engineering role.
I've added the data analyst role recently because machine learning has become so difficult to break into. If you're new to IT, I'd highly recommend this path. You can be in a real-world data analsyt role in under a year if you really apply yourself. LogikBot is the most affordable and fastest way to attain this role. The platform even has an exam simulator for preparing for the exam.
For the data analyst role, you take two courses then prepare for and pass exam PL-300. This exam, PL-300, is the Microsoft Certified PowerBI Data Analyst Certification.
If you're in a shortcut role, then there's an entire path laid out for learning machine learning. The platform used by Uber and Microsoft. I worked at both companies. The courses are real-world and comprehensive. This is not the easy way. There is no easy way. This is one of the few platform that's offers a real-world approach to working in AI.
I finally found the BS filter for the tech industry. Haha, great content. Short and straight to the point, always.
Thank you. Appreciate it. Thanks for watching.
On god. 😂
This is kind of a relief haha, this doesn't seem too bad I thought it was going to get crazy like the later projects in my DB classes at school
This is exactly what I was just wondering.
Perfect timing.
Thanks!!!
wow loved this video very informative!!!
Thanks. Thanks for watching!!
Pretty Straight forward...
I like the way he speaks...
Thanks.
Hey Mike. I learned statistics, design of experiment etc. at university. Also, completed my MSc by writing a paper about to predict some psychological score by using sensor data and built a basic classification model for that. I cleaned data, made some insights about it by using Python and Excel. Data was not so complicated so Excel was enough for me. As I am not a software engineer or cse, I have not needed SQL yet. Now I am looking for PhD positions in Europe, and I want to build more complicated models and study them at PhD. I mean using data from smart watches (time-series based), application using data (like instagram photo zooming behavior, daily heart rate, step number etc.) and then create more complicated model to estimate behaviours. Which path should I follow? I checked logikbot and saw that there are lots of contents but I could not decide where to start, and how to continue after that. A little insight from an experienced person in that field like you would be great for me. Thanks for the great platform you create in advance.
It looks like you're on the path to becoming a researcher. Is that what you want? If you want to do research you are headed in the right direction. You'll need to be in a PhD program before applying to reseach roles. Once you're in a PhD program, companies will take you seriously for research roles.
If you don't want to do research and you want to work in the real-world as a MLE, you need to get a job. :) This might be difficult because of you don't have SQL but it sounds like you know enough for an entry level role and might be able to talk you're way into a role. Your first job is really important. Experience in the real-world is the number one job criterion. If you do decide to take courses on my platform, start with SQL, then data cleansing and then move on to dealing with imbalanced data. Most real-world machine learning is classification and the data is dirty and unbalanced.
Like your videos , straight to the point. I have a masters in applied mathematics and computer science and I am very interested in starting a career in machine learning. My university offers a masters program in ML. What route should I take. Should I start as a data analyst?
You should get a job. A master's in this space is worthless. I'd suggest the DA route.
@@thedatajanitor9537 how is a master in applied mathematics and csi or ML useless?
I have worked with some 'Data Analysts' in paying jobs who don't know SQL. They rely on those who do to create models in Power BI and SAP Universes (for example) to give them a framework to extract data into Excel, or to author reports. I have never worked on Wall St but I have enough finance/capital markets experience to know that there are several roles there that are purely Excel based. I have also heard of Excel gurus building entire trading systems in Excel. I have also heard of the mythical Excel geniuses who apparently are so proficient in Excel they solely use the keyboard and consider anyone using a mouse in Excel to be 'sub-par'. I wonder how true that is...
Bottomline is that a decent proficiency in SQL is crucial. Excel is a good start but eventually you need to transition into more powerful enterprise tools like PowerBI. Also much of the data in the REAL world is structured. Even the so called unstructured data has to be structured and stored in RELATIONAL databases before any meaningful analysis. You can connect to a database and perform some magic but there are limitations or should i say lack of flexibility and power. Back to SQL again. That is why Mike is preaching SQL + a DATA tool like POWERBI as key skills before you attempt ML, DE etc...Excel has its place but its a TOY tool when it comes to ML
@@apamwamba I agree - a decent proficiency in SQL is crucial.. preferably T-SQL. Excel is a second rate tool when used as a data store but it is an amazingly powerful, if abused, data analysis tool in the hands of an expert. Every day hundreds of millions - no, billions - of dollars of investment decisions are made based on analysis in that product. I hate cleaning up Excel macros as much as the next person and I hate dealing with Excel but it is here to stay.
@@joshash5944 Its a good tool for the right people working on relatively small datasets, esp data analysts and other domain experts. A few minutes ago,I was trying to do something with power query and it crashed. I have had to do the task in PowerBI and export to csv. Excel has its solid place but not in the high end ML arena
@@apamwamba Building enterprise Power BI datasets is one of the things I do for a living... we don't use Power Query for transformations and we don't build models in Power BI Desktop...
@@joshash5944 We seem to be in agreement that these tools have their place. Truth is that Excel is not for high end ML work.
Hey mike! Just quick question do you think learning sql from books is an effective way? The old school style and How do you learn stuff?
I learn via courses and books. Both are great. It doesn't matter as long as you are learning the right stuff.
@@thedatajanitor9537 one more question mike! So ive been reading comments people usually taking ms programs .
I think if they wana take somehow ..it will be better if its ms in data analytics right? (Atleast 10% good?) instead completely diff.
It won't make a difference. They are all theory and zero practical application. If you forced me to get an MS it would be in machine learning but that would be a waste of years of my life.
@@thedatajanitor9537 agreed!
@@thedatajanitor9537 Hey Mike. I started with basic MySQL for DA course last month then i realized i need an interesting dataset to practise upon. Having tried many sites, i decided to scrape IMDB site and make my own database. Basically, it is movies(5600 noteworthy ones ),directors and actors table connected. Then i proceeded to designing & solving dozens of difficult tasks to comprehend sorta advanced concepts. And thanks to w3schools documentation and ChatGPT, now i am able to solve those problems on my own without assistance. I can share it with you if you're interested.