Photo by author
# Introduction
Data science is often confused with machine learning, but it is actually much more than that. It’s about collecting, cleaning, analyzing and visualizing data to find useful patterns that can help us make decisions. Machine learning is only a small part of this big picture. I started this series of fun projects to encourage hands-on learning because honestly, you don’t learn data science by watching endless theory. You learn by building it.
For this article, I’ve chosen five projects that cover different stages of a basic data science workflow, from basic data cleaning to data exploration, to building models, and even deploying them for real-world use.
# 1. The only data cleaning framework you need
This video It’s from Christine Jiang, who works as a data analyst, and she shares a really practical approach to data cleaning that I think anyone working on projects will find useful. When cleaning data, we often think about “how clean is too clean,” and Christine shows a clear way to handle this using her five-step clean framework. She walks through how to solve unsolvable problems, standardize values, standardize everything, and aim for “perfect” to make your data reliable. The examples she uses, like fixing missing country codes or inconsistent product descriptions, are very relatable and the mindset she emphasizes is just as important as the tools. I found this to be a must for anyone trying to efficiently manage real-world data.
# 2. Analysis of search data in pandas
This video It shows why simply having data isn’t enough and how looking carefully at the numbers can reveal hidden patterns. The presenter walks through the datasets, summarizing distributions, checking for missing values ​​and outliers, and looking at relationships between columns. Pandas And Seaborne. I found it really practical because it doesn’t just show the orders, it explains why each step matters and how statistics can tell you things that aren’t obvious at first glance. It’s a great guide for anyone who wants to explore real-world data and gain meaningful insights before jumping into modeling.
# 3. Visualization of data using pandas and plots
This video By Greg Komdt, founder of Data Independent, it shows that telling a story with your data is just as important as building models. He walks through a hands-on tutorial using pandas and for data wrangling From the plot For interactive charts, starting with the basics of what makes visualization effective. You’ll see how to load and format data, choose the right chart types, and add formatting touches that make your charts clear and easy to understand. I really liked how practical it is, with tips on real-world issues like outliers, date axes, and aggregations, and how small choices can improve readability. Finally, you’ll learn how to create interactive, shareable charts that effectively communicate insights.
# 4. Characterization of engineering techniques for machine learning in Python
Once your data is clean and understandable, it’s time to create better features. This lesson Focuses on the “Feature Engineering” phase, where you transform and generate new data columns that can improve your model. The instructor explains techniques such as encoding categorical variables, handling missing data, dimensionality reduction (principal component analysis (PCA)), and generating interaction terms. I like that it also highlights what not to do about leaking data, more proper and more engineering features. It’s a great resource for anyone who wants to go from raw data to building well-engineered features for real-world machine learning.
# 5. Deploying a machine learning model in the Streamlet app and making live predictions
Finally, the most satisfying part – bringing your model to life. In this tutorialYiannis Pitslides shows how to deploy using a trained machine learning model Streamlet. He walks through loading a stored model, setting up a clean interface with input boxes and buttons, and generating real-time forecasts for car prices. The video even includes a feature importance concept Plotlyso you can see which inputs are given the most importance. I liked how practical it was, with tips on keeping raw and cleaned data separate, handling dependencies, and running the app locally or on a host. It’s a short tutorial, but it does the job beautifully and gives you that “end-to-end” experience that most beginners miss.
# wrap up
These projects cover all the key steps in the data science workflow and demonstrate how theory comes to life in practice. Grab your datasets and start experimenting. There is no better way to learn data science.
Kanwal Mehreen is a machine learning engineer and technical writer with a deep passion for data science and the intersection of AI with medicine. He co-authored the eBook “Maximizing Productivity with ChatGPT”. As a 2022 Google Generation Scholar for APAC, she champions diversity and academic excellence. He has also been recognized as a Teradata Diversity in Tech Scholar, a MITACS GlobalLink Research Scholar, and a Harvard Wicked Scholar. Kanwal is a passionate advocate for change, having founded the Fame Code to empower women in stem fields.