Last Updated on August 2, 2023 by Editorial Team
Originally published on Towards AI.
How to Be a Great Data Scientist and Not a Garbage One: Tips and Tricks to kick off Data journey in the right way
“How can I become Data Scientist within one week ?”
“Is there any capsule content to learn AI ML just within one day?”
“Any Moocs to learn Analytics overnight to earn 6 figures next week?”
These are the most common question I use to get on my plate from social media from many AI and Analytics enthusiasts every day.
Daily tons of newsletters, MOOCs, and Courses are getting launched by Non-AI/ML Influencers to make learners fools by showing ridiculous dreams and injecting many worthless imaginations and practices.
I have seen tons of recently became DS or AI guys not getting jobs with minimum wages, which can help them even to survive after competition of their DS University courses or Inlfunecers Moocs capsule.
Let’s understand why?
Googling: “What is Data Science?”
Yes, Googling is very important all the time; even before kicking off something absolutely.
It does not matter what subject you are about to learn. Learners need to align 3 things before kicking off their learning path:
- Whether the learner’s background is the same or different.
- Understanding the prerequisites of the subjects before diving in.
- How to apply after acquiring the knowledge in the real world?
Having a clear idea about the subject before starting to learn increases the success of learning.
It doesn’t matter whether you search on Google or ask Bard or Chatgpt. Using these tools, you can navigate your learning journey.
This simple applied to any subjects.
This philosophy is very essential and mostly followed by every successful expert in every field in this world.
Either you are about to start your career in the Data Science field or you are planning to switch from another background to this field. You should have a clear reason behind kicking it off, just buying a couple of courses or joining training classes.
A couple of my friends and ex-colleagues changed in this domain from different fields of expertise. But, they did not become Data scientists; neither they wanted to. They learned analytics as a tool that serves their individual purposes.
Tools are tools:
Learning Python or R or any other language or ML tools won't make you a data scientist. These are just tools to be utilized by ML practitioners to solve the business problems that need to deliver at the customer’s end or might be for personnel research.
The day after tomorrow a tool may come and replace them like a cakewalk. But, that’s not gonna change the shape of Analytics. They will change the tech stack of the problem that you will be solving then; not your solution.
Domain Selection and making a roadmap
Selecting the specific domain and creating a roadmap in that?
If, you are an aerospace engineer and you are learning Data science and AI techniques. You can focus on what kind of equipment you are dealing with every day, and then you can learn only that part of Data science.
Suppose you are learning DS to apply in a turbine power generation system. Simply learn the domain well where AI can add some value and then learn only the relevant tools and techniques to solve them. For example, you are trying to detect faults in advance in the turbine and about to forecast the power generation. Then, you need to learn Time series techniques and details of sensor data and dedicated forecasting and Anomaly detection algorithms only. You may not need to spend time in NLP at that point.
“You have no need to learn everything. Never!”
Non-IT to IT trend:
You won’t have to jump into IT after learning DS or AI. You can apply this knowledge in any field.
After, selecting the domain to apply and selecting tools and techniques to learn; it will be easier to create a complete solid roadmap.
Nowadays, we see everyone choose to go to the IT field to mine 6 figures at the end of the month. But, this is simply camouflaged to squeeze the pocket of learners by Ed-etch companies and individuals.
You can always earn more than 6 figs if you can really solve real life with just a simple hammer.
If, everyone is learning LLM and Chatgpt and making millions overnight using Prompt engineering kinda stuff around you. You don’t need to be reactive by jumping with them in the same. These are just noisy and hype trends.
You will just end up with hype nothing else; which will barely help you to solve business cases in real-time during your work.
Discard all kinds of hype during your learning journey.
Still, now, you can learn from this Old Gold: Link
Consult with Experts who have been working in the field for longer; instead of social media influencers. Many Social media Influencers have made AI and DS courses and are selling and earning million. On the other hand, learners are draining their pockets with zero output.
Find someone who has been in this industry and contributed to many projects with his or her competency (timestamped evidence either on GitHub or LinkedIn or anywhere). Consult with them about learning strategy and what you are planning to achieve; even if you are learning from open source content. That’s fine.
Learning in a conventional way:
Best way to do this is by learning the 3-way technique.
- Learning Statistics and ML: Complete theory with detailed concepts
2. Learning Tools as required (for example — Python, Rapidminer, Julia, etc.)
3. Learning representation: After developing a complete AI ML strategy. How will you represent that to the business stakeholders, partners and peers, and your boss?
Representation of your work in Data science matters a lot. Building fancy models which can not be perceived by stakeholders or consumers ends up in a mess.
I will always suggest learning in a conventional way. Instead of trying to learn from ticktock or Instagram influencers who does daily marketing vlog.
Learning 20 topics in 20 minutes is nothing more than just garbage.
Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.
Published via Towards AI