.
“Companies are investing heavily in data science”.
After completing my engineering and starting my job, I was continuously bombarded with these statements on the Internet. I was puzzled and, how Lord Buddha wanted to know the truth of life, I also wanted to clarify my doubts. To search for answers, I searched the internet and reached out to many people inside and outside of this domain.
In this article, I have compiled how I decided to start my journey to data science. We will discuss how to build your digital profile and
Is data science real or just old wine in a new bottle?
Data science was always present among us. Excel, SQL, Statistics are the tools of data science for the early age. This does not make this field obsolete. Data science will always surprise us with new and updated magic tricks. Previously, we used to input data into an excel sheet and then plot charts. Today, data is automatically stored and charts are plotted automatically with advanced visualization tools. With the advance, data science has given us a lot of buzzwords like machine learning, deep learning, AIOps, etc. and will continue to do so.
There is nothing like a one size fits all
To be honest, there is no definite path and there shouldn't be one either. The true essence of data science is people of different backgrounds and technicalities working together. From any trip I've covered, I can just give you course names and tell you to complete those courses and do those projects, but this will feel like a burden to you. You will be in a rat race completing courses and projects and, Finally, will run out. I am an electronics engineer whose year-end project was the Raspberry Pi facial recognition smart door, whose interest led him to learn Machine Learning and Deep Learning. I also did not follow a specific path, I always went with what I liked and it kept me awake at night.
Learning data science: Yes or no?
Many of you might be wondering if a non-computer person / computer science can learn data science or not. The answer is yes. A person other than CS / IT can learn data science and is not an obligation for CS / IT learn data science.
No need to start DS, ML or AI due to peer pressure. You may feel left out if you aren't preparing extensively for the wrath of the AI. If you are good at what you are doing and you love what you are doing, so it's amazing to keep doing what you're doing and keep up to date by reading news and blogs.
Like biodiversity it is necessary to balance the ecosystem, similarly, technological diversity is necessary for a thriving community. We will always need mechanical engineers, electrical, artists, web developers, app developers, content creators, doctors, filmmakers, THAT, athletes, etc.
How Deep Should You Dive Into Data Science?
Then, Let's take a look at some domains associated with data science that can be followed:
- Data engineering and data warehousing: takes care of storing and querying the data for future analysis. Keeping data is as important as making predictions. A good prediction model requires good quality data.
- Cloud and Distributed Computing: if you are an IT networking expert, become familiar with the data science project life cycle and you will be able to design and implement data science models for easy access.
- Intelligence and business strategy: if you are an expert in the domain, then it is the backbone of the entire data science project. A BI strategist is responsible for managing dashboards, inform stakeholders, test and validate models and document.
- IoT developers: if you like hardware and you like to build circuits and controllers, can play the role of collecting the data by sensors and preparing it for analysis or storage in real time.
- Computer vision: if you love image processing, can apply deep learning concepts and work on process automation and object detection model building.
- ML Engineer: machine learning engineers feed data into models defined by data scientists. They are also responsible for taking theoretical data science models and helping scale them to production-level models that can handle terabytes of data in real time..
- NLP Engineers: NLP engineers responsibilities include transforming natural language data into useful functions using NLP techniques to feed classification algorithms.
In the future, there will be many more new job profiles coming onto the scene, the only thing that will keep you ready for work is continuous learning.
Then, What's my guide on machine learning?
I enrolled in various courses at my own pace. I did not restrict myself with the videos and the tasks that the courses gave me, but after a while I used to read the name of the topic and started to learn from the research papers, internet searches, books and other sources. You can also select any course that feels good and fits in your pocket. See how content is delivered and what other services are offered. I will share with you my roadmap on how I prepared. I won't limit you to the rigid timeline, you can also follow it at your own pace.
Mathematics and Statistics
Learn the basics of statistics and brush up on your school and university math.
- Derivative and function minimization
- Vector y matrices
- Probability distribution
- Random variables
- Normal distribution
- Z score
- Hypothesis testing
- Z test and T test
- Chi-square test
- ANOVA test
- Statistics F
Only basic knowledge is enough of the above topics.
Basic programming
You can select any language of your choice. I chose Python. Some fundamental concepts to know:
- Type of data
- Terms
- Loops
- Features
- Object-oriented programming
- Exceptions handling
Don't worry if you are not comfortable with these concepts at first.. After a lot of practice, will be sure.
Machine learning
Understand how the different algorithms work and how to implement them:
- Linear regression
- Logistic regression
- Decision tree
- Random forest
- Set, bagged, impulse
- XGBoost
- Bayes ingenuo
- KNN
- Grouping of K-stockings
- Hierarchical grouping
- DBSCAN
- Principal component analysis
- Support vector machines
- Time series and anomaly detection
Seasoned professionals can start with advanced Excel and other data visualization tools such as PowerBI and Tableau. You can get many platforms where you can make ML predictions without coding knowledge.
If it's cooler, start learning programming and complete some ML courses. Experiment a lot and keep working hard. Motivational word to refresh: “If ML is your passion, then i know like batman: do your office work during the day and follow your passion at night.
If you are in university, attend conferences, workshops, tech festivals, complete courses, participate in hackathons, meet as many people as you can and, the most important, enjoy the process.
Retaining what you learn
Now that you are in learning, let me share with you some tips on how to retain what you learn:
- Practical implementation of the project. Start with a basic project and expand it. Integrate it with apps, deploy it on cloud platforms, etc.
- Feynmann learning technique:
- Choose a concept you want to learn about
- Imagine you are teaching it to a sixth grader.
- Identify the gaps in your explanation; Go back to the original material to understand it better.
- Review and simplify
- You can record videos of yourself explaining concepts and upload them to YouTube. In the end, will have your own video notes to refer to.
No solo YouTube, you can choose any social network, how I save my digital notes on Instagram and Facebook.
https://www.instagram.com/pandaspython/
https://www.facebook.com/pandaspython08
Building your digital profile will not only help you retain things, it will also help you a lot in making connections with others in the industry. You can show your work, collaborate with others and work on projects. This practice will develop communication and general personality skill that many people lack..
Add each and every bit of code you make on Github and write a nice readme.md about it. Start blogging or create a website to show what you are doing (google sites are also sufficient). Simply digitally record each and every one of your activities. Blogger, medium facebook page that you can write anywhere you want.
Final notes
So this was my article on what data science is and how far one should learn. I have also shown how to learn effectively. If you have questions and want suggestions on which books or blogs to follow, ask me in the comment section. Finally, I would like to tell you to enjoy the process rather than just the courses.
