How to be a 10X Data Scientist?
Be an elite, be an extraordinary not be an average.
For over 6+ years, I have worked with many people in my Data Science career. 99% of them are average; however, 1% of them were making a difference.
In this article, I will share with you my experience with those people and provide you with the tips that I observed, which may help you become one of them. And yes, this article, of course, includes ChatGPT knowledge.
Let’s get started.
1. Problem-Solving Attitude
It is a cliché, but it is an inevitable skill in the job description, especially in Data Science. To achieve data science tasks, you must follow specific steps, right? Starting with Data Exploration to Machine Learning.
The top 1% professionals I’ve worked with share this typical attitude; they approach challenges with a mindset geared towards finding solutions efficiently. They break down the problem into smaller pieces and solve it as simply as possible.
Example: Data Cleaning Challenge
Average Approach: An average Data Scientist might begin manually identifying and correcting errors, spending days or even weeks on the task.
1% Approach: A professional with a problem-solving attitude would, instead, evaluate different automated cleaning tools. Because that would be the first step to make the process basic and accurate, and they will select the most suitable one, completing the task in a matter of hours.
The difference here is not just in speed, but also in the ability to tackle more complex challenges by dividing them into smaller pieces and solving them with basic methods.
By thinking critically and choosing the right tools, especially in this era of AI tools, you will boost your skills, and they will perceive you as a team instead of a one-person operation.
2. Basic ChatGPT Knowledge is not enough now!
If you want to find a more efficient way to solve any task, consider using ChatGPT.
Why?
With the release of ChatGPT, LLMs have transformed many fields, including Data Science. Using your power in the task now will differentiate you from a one-time Data Scientist.
By using ChatGPT now, you can easily enhance the following stages of data science.
Data Scraping
Data Exploration &Data Analysis
Data Visualization
Machine Learning
But how? Let me provide an example: you are going to conduct data analysis with ChatGPT.
Of course, you can do it by using Python, but why bother now? Of course, I advise you to review the code ChatGPT provided, but you can easily automate Data Analysis with the Noteable plug-in.
By just using a simple prompt, you can initiate Data Analysis.
Here is a simple prompt that will help you:
Load this dataset : "Link"
Use this as my default project: "After signing noteable, in the website you can find this one's link, go to projects"
Act as a data scientist and analyze this dataset.Then it will update your notebook according to your requests in ChatGPT. That's easy.
How can you write code faster than this?
3. Programming is a must!
In Data Science, despite the increasing number of LLMs like ChatGPT, Claude 2, that can code for you, having basic coding knowledge is not optional—it's a necessity.
Why is programming essential? Here's why:
Automation: With programming skills, you can automate repetitive tasks and have a lot of free time to enhance the quality of your work.
Customization: Tools and plugins can only take you so far. Knowing how to code lets you find solutions to specific problems. You can not copy-paste everything from ChatGPT or other LLM tools; you should know what that means.
Example: Automate Data Collection and Cleaning
Without Programming Knowledge: Let's say you're tasked with collecting and cleaning data from various online sources for an analysis project. Without good programming skills, you might have to manually browse websites, copy data into spreadsheets, and clean inconsistencies. This process could be extremely time-consuming, error-prone, and inefficient, potentially taking weeks to complete.
With Programming Knowledge: Now, imagine you have skills in programming and familiarity with services like AWS Lambda. You could write a script to automatically scrape the data from the web, then use AWS Lambda to set up an automated pipeline that regularly fetches, cleans, and stores the data, ready for analysis.
While tools like ChatGPT have revolutionized the way we work, a solid understanding of programming remains a vital skill.
4. Statistics are not boring!
When you look at the data, what do you think? What will you look at first? The Data Scientist, who also wore the suit of Statistician, first thinks:
What is the shape of this dataset
Are there any outliers that I should address?
Is the data skewed? If so, which method should I use to normalize my data?
However, the 1X data scientist follows the checklist from the course he took. If a critical case arises, such as skewed data that a statistician should recognize, proper care should be taken before applying the ML algorithm.
I understand that people who dislike statistics often have a valid point, and in my opinion, the main reason is that the instructors were not engaging. There are fantastic ways to learn statistics.
Let’s say you are a sports fan and a Fenerbahçe fan. Your team has a match next Saturday, and after reading the article above last week, you planned to apply your knowledge to real-life events. How would you calculate?
With the Poisson distribution, you can do this calculation. Do you know that the probability of two people having the same birthday in a group of 23 people is approximately 50%?
Final Thoughts
If you are reading so far and saw my thumbnail, either you first think, Who is the man in the thumbnail picture or or you know who he is. If Nikola Tesla were alive in this era, he would be a 10x Data Scientist for sure.
Be innovative, utilize AI tools, but don’t let them use you instead; take control and avoid copying and pasting everything from them.
It is the era of AI tools for sure, but I think the one who will survive is the one who becomes a master of these tools and uses them wisely, where they are needed, not everywhere.
Thanks for reading.
“Machine learning is the last invention that humanity will ever need to make.”
Nick Bostrom