12 Data Science Projects To Try (From Beginner to Advanced)

From breast cancer detection to user experience design, businesses across the globe are leveraging data science to solve a wide range of problems. Every mobile/web-based product or digital experience today demands the application of data science for personalization, customer experience, and so on. This opens up a world of opportunities for data science professionals.

To land a data science job, however, early career professionals need more than just a strong theoretical foundation. Hiring managers today are looking for data scientists who have the hands-on experience of delivering projects that solve real-world problems. Even before you land your first job, you need to have ‘experience’ demonstrating your ability to deliver them. No sweat. We’ve brought help.

What Is a Data Science Project?

A data science project is a practical application of your skills. A typical project allows you to use skills in data collection, cleaning, analysis, visualization, programming, machine learning, and so on. It helps you take your skills to solve real-world problems. On successful completion, you can also add this to your portfolio to show your skills to potential employers.

Data Science Projects To Try

Whether you’re a complete beginner or one with advanced skills, you can gain hands-on experience by trying out projects on your own or working with peers. To help you get started, we’ve curated a list of the top 15 interesting data science projects to try. See what catches your fancy and get started!

Beginner Data Science Projects

“Eat, Rate, Love”—An Exploration of R, Yelp, and the Search for Good Indian Food

When it comes time to eat, many people turn to Yelp to choose the best options for the type of food they’re looking for. They search, eat, rate, and leave reviews for the restaurants they’ve visited. This makes Yelp a great source of data to run data science projects. 

A Springboard Data Science Bootcamp graduate Robert Chen chose this data to explore if the best reviews led to the best Indian restaurants. Chen discovered while searching Yelp that there were many recommended Indian restaurants with similar scores. Certainly, not all the reviewers had the same knowledge of this cuisine, right? With this in mind, he took into consideration the following:

• The number of restaurant reviews by a single person of a particular cuisine (in this case, Indian food). He was able to justify this parameter by looking at reviewers of other cuisines, such as Chinese food.
• The apparent ethnicity of the reviewer in question. If the reviewer had an Indian name, he could infer that they might be of Indian ethnicity, and therefore more familiar with what constituted good Indian food.
• He used Python and R programming languages.

His modification to the data and the variables showed that those with Indian names tended to give good reviews to only one restaurant per city out of the 11 cities he analyzed, thus providing a clear choice per city for restaurant patrons.

Yelp’s data has become popular among newcomers to data science. You can access it here. Find out more about Robert’s project here.

Customer Segmentation with R, PCA, and K-Means Clustering

Marketers perform complex segmentation across demographic, psychographic, behavioral, and preference data for each customer to deliver personalized products and services. To do this at scale, they leverage data science techniques like supervised learning.

Data scientist Rebecca Yiu’s project on market segmentation for a fictional organization, using R, principal component analysis (PCA), and K-means clustering, is an excellent example of this. She uses data science techniques to identify the prospective customer base and applies clustering algorithms to group them. She classifies customers into clusters based on age, gender, region, interests, etc. This data can then be used for targeted advertising, email campaigns, and social media posts. 

You can learn more about her data science project here.

Road Lane Line Detection

To follow lane discipline, self-driving cars need to detect the lane line. Data science and machine learning can play a crucial role in making this happen. Using computer vision techniques, you can build an application to autonomously identify track lines from continuous video frames or image inputs. Data scientists typically use OpenCV library, NumPy, Hough Transform, Spacial Convolutional Neural Networks (CNN), etc., to achieve this.

You can access a sample video for this project from this git repository here.

Intermediate Data Science Projects

NFL Third and Goal Behavior

The intersection of sports and data is full of opportunities for aspiring data scientists. Divya Parmar, a lover of both, decided to focus on the NFL for his capstone project during Springboard’s Introduction to Data Science course. His goal was to determine the efficiency of various offensive plays in different tactical situations. 

Parmar collected play-by-play data from Armchair Analysis, and used R and RStudio for analysis. He developed a new data frame and used conventional NFL definitions. Through this project, he learned to:

1. Assess the problem
2. Manipulate data 
3. Deliver actionable insights to stakeholders

You can access the dataset here. 

Who’s a Good Dog? Identifying Dog Breeds Using Neural Networks

Image classification is one of the most popular and widely in-demand data science projects. Classifying dogs based on their breeds by looking into their image is a highly loved data science project. Garrick Chu, a graduate of Springboard’s Data Science Career Track, chose this for his final year submission. 

One of Garrick’s goals was to determine whether he could build a model that would be better than humans at identifying a dog’s breed from an image. Because this was a learning task with no benchmark for human accuracy, once Garrick optimized the network to his satisfaction, he went on to conduct original survey research to make a meaningful comparison.

He worked with large data sets to effectively process images (rather than traditional data structures) with network design and tuning, avoiding over-fitting, transfer learning (combining neural nets trained on different data sets), and performing exploratory data analysis. 

To do this, he leveraged neural networks with Keras through Jupyter notebooks. You can explore more of Garrick’s work here and access the data set he used here.

Uber’s Pickup Analysis

Is Uber Making NYC Rush-Hour Traffic Worse?—This was one of the four questions answered by FiveThirtyEight, a data-driven news website now owned by ABC. If you are looking to improve your data analysis and data visualization skills, this is a great data science project. 

For this, FiveThirtyEight obtained Uber’s rideshare data and analyzed it to understand ridership patterns, how it interacts with public transport, and how it affects taxis. They then wrote detailed news stories supported by this data analysis. You can read their work of data journalism here. You can access the original data on Github.

Predicting Restaurant Success

Here is another Yelp-based project, but more complex than the one we discussed earlier. Data scientist Michail Alifierakis used Yelp data to build his “Restaurant Success Model” to evaluate the success/failure rates of restaurants. He uses a linear logistic regression model for its simplicity and interpretability, optimized for the precision of open restaurants using grid search with cross-validation.

This is a great data science use case for lenders and investors, helping them make profitable financial decisions. You can learn more about the project from here and take a look at the code on GitHub.

Predictive Policing

Many law enforcement agencies worldwide are moving towards data-driven approaches to forecasting and preventing crimes. They leverage data science technologies to automate the pattern detection process that will help to reduce the burden on crime analysts. Data scientist Orlando Torres launched a data science project on predictive policing, albeit to unexpected results. He used data from the open data initiative and trained the model on 2016 data to predict the crime incidents in a given zip code, day, and time in 2017. He used linear regression, random forest regressor, K-nearest neighbors, XGBoost, and deep learning model — multilayer perceptron.

With this data science project, he learned that it is very easy to lose explainability while building models. He writes, “if we start sending more police to the areas where we predict more crime, the police will find crime. However, if we start sending more police anywhere, they will also find more crime. This is simply a result of having more police in any given area trying to find crime.” Given the number of law enforcement agencies using data science for policing, it almost feels like a self-fulfilling prophecy.

You can read more about his project here.

Building Chatbots

Today, businesses are automating their customer services with chatbots. Creating your own chatbot can be a great data science project too. The two types of chatbots available today are domain-specific chatbots and open-domain chatbots. They both use Natural Language Processing (NLP) and Recurrent Neural Networks (RNN). For an intermediary data scientist, you can perhaps take this up a notch—try creating a sensitive chatbot with capabilities to detect user sentiment.

Patrick Meyer runs a data science project of this kind. He discusses using the polarity system to identify happy, neutral, and unhappy; Paul Ekman’s initial model with six emotions—anger, disgust, fear, joy, sadness, and surprise or his extended list of sixteen; Robert Plutchik’s wheel of emotions and Ortony, Clore, and Collins (OCC) model. 

You can learn more about his detection techniques here. And access the dataset here.

Advanced Data Science Projects

Amazon vs. eBay Analysis

Finding the lowest price for a product on the Internet makes up a significant part of online shopping. Chase Roberts decided to make that easier. In support of a Chrome extension he was building, Roberts compared the prices of 3,500 products on eBay and Amazon. The results showed the potential for substantial savings. For his project, Roberts built a shopping cart with 3,520 products to compare prices on eBay vs. Amazon. Here’s what he found:

• If you chose the wrong platform to buy each of these items (by always shopping at whichever site has a more expensive price), this cart would cost you $193,498.45. (Or you could pay off your mortgage.) This is the worst-case scenario for the shopping cart.
• The best-case scenario for our shopping cart, assuming you found the lowest price between eBay and Amazon on every item, is $149,650.94. This is a $44,000 difference—or 23%!

You can read more about his project, starting with how he gathered the data and documenting the challenges he faced during this process.

Fake News Detection

A recent study revealed that false news spread faster and reached more people than the truth and around 52% of Americans shared that they regularly encountered fake news online. A four-person team from the University of California at Berkeley built a fake news classifier. For this, the team focussed on clickbait and propaganda, the two common forms of fake news. They then developed a classifier that would detect these two forms. Their process involved: 

• Taking data from news sources listed on OpenSources
• Used NLP to do the preliminary processing of articles for content-based classification
• Trained various machine learning models to divide the news articles
• Developed a web application to act as the front end of their classifier. 

You can learn and try out more about this here.

Audio Snowflake

When you think about interesting data science projects, chances are you think about how to solve a particular problem, as seen in the examples above. But what about creating a project for the sheer beauty of the data? For her Hackbright Academy project, Wendy Dherin did just that. 

She developed Audio Snowflake to create a splendid visual representation of music as it played, capturing specific components like tempo, key, mood, and duration. Audio Snowflake mapped both quantitative and qualitative characteristics of songs to visual traits like saturation, color, rotation speed, and figures it produces. 

Read more on this project here.

Visualizing Climate Change

2020 was recorded as the warmest year to date by NASA, and the last seven years have been the warmest seven years on record. Climate change is one of the most pressing issues humans face today. It is more important than ever to spread awareness and inform people of the magnitude of this problem. Data visualization can play a crucial role in that. 

The data scientist Giannis Tolios did a project where he visualized the changes in global mean temperatures and the rise of CO2 levels in the atmosphere using Python. He uses various libraries such as Pandas, Matplotlib, and Seaborn for the data, visualizing it in line graphs and scatterplots. If climate change is a topic you want to work on, you can learn more about the project here.

Democratizing Data Science at Uber

One of the key challenges in data science is that it requires one to be a mathematician or a statistician even to make basic predictions and forecasts. Uber’s data science platform overcomes this challenge by automating forecasting using pre-built algorithms and tools, enabling everyone on the team to get predictions as long as they have data. 

Director of Data Science at Uber, Franziska Bell, talks about how they plan to give the capabilities of a data scientist to every Uber employee. This way, Uber uses artificial intelligence, machine learning, and data science to solve real-world problems. Read more about it here.

Credit Card Fraud Detection

With online and digital transactions gaining more popularity today, their chances of being fraudulent are also on the rise. Therefore banks and financial institutions are looking to leverage data science techniques to identify fraudulent transactions and prevent them from being executed. By processing data across customer location, behavior, transaction value, network, payment method, etc., you can train the algorithm to detect anomalies. You can build your classification engine for fraud detection using decision trees, K-nearest neighbor, logistic regression, support vector machine, random forest, and XGBoost.

To get started, you can find datasets here.

Datasets for Data Science Project Ideas

Here are some online data sources which you can access and download for free for your data science projects:  

 VoxCeleb. A gender-balanced, audio-visual data set containing short clips of human speech from speakers of different ages, professions, accents, etc. They are extracted from interviews uploaded to YouTube. It can be used for various applications like speech separation, speaker identification, emotion recognition, etc.

 Boston Housing Data. A fairly small data set based on the information collected by the U.S. Census Bureau data regarding housing in Boston. This data set can be used for assessment, focusing on the regression problem.

Kaggle. With over 50,000 public datasets on a wide range of topics, you can find all the data and code that you require to do your data science project ideas. They also offer competitive data sets that are clean, detailed, and curated. 

National Centres for Environmental Information. The largest storehouse of environmental data in the world, this provides information on the oceanic, atmospheric, meteorological, geophysical, climatic conditions, and more. 

Global Health Observatory. If you are interested in doing projects in the health industry, then this is the best place to get the data you need. It also has some of the latest COVID-19 data. 

Google Cloud Public Datasets. A place where you can access data sets that are hosted by  BigQuery, Cloud Storage, Earth Engine, and other Google Cloud services. 

Amazon Web Services Open Data Registry. This has an extensive repository of data sets that you can either download and use or analyze on the Amazon Elastic Compute Cloud (Amazon EC2). You need to first create a free AWS account to get access to the data sets. 

Tips for Creating Interesting Data Science Projects

To help you navigate the world of data science projects, we asked Springboard mentors and instructors for their advice. Here’s what they had to say. 

Choose the Right Problem

If you’re a data science beginner, it’s best to consider problems that have limited data and variables. Otherwise, your project may get too complex too quickly, potentially deterring you from moving forward. Choose one of the data sets in this post, or look for something in real life that has a limited data set. Data wrangling can be tedious work, so it’s critical, especially when starting out, to make sure the data you’re manipulating and the larger topic is interesting to you. These are challenging projects, but they should be fun!

Breaking Up the Project Into Manageable Pieces

Your next task is to outline the steps you’ll need to take in order to create your data science project. Once you have your outline, you can tackle the problem and develop a model to prove your hypothesis. You can do this in six steps:

• Generate your hypotheses
• Study the data
• Clean the data
• Engineer the features
• Create predictive models
• Communicate your results

Generate Your Hypotheses

After you have your problem, you need to create at least one hypothesis to help solve the problem. The hypothesis is your belief about how the data reacts to certain variables. 

This is, of course, dependent on you obtaining the general demographics of specific neighborhoods. You will need to create as many hypotheses as you need to solve the problem.

Study the Data

Your hypotheses need to have data that will allow you to prove or disprove them. Look in the data set for variables that affect the problem. If you do not have the data, either dig deeper or change your hypothesis.

Clean the Data

As much as data scientists prefer to have clean, ready-to-go data, the reality is seldom neat or orderly. You may have outlier data that you can’t readily explain, like a sudden large, one-time purchase of an expensive item in a store that is in a lower-income neighborhood. Or maybe one store didn’t report data for a week.

These are all problems with the data that aren’t the norm. In these cases, it’s up to you as a data scientist to remove those outliers and add missing data so that the data is more or less consistent. Without these changes, your results will become skewed, and the outlier data will affect the results, sometimes drastically.

Engineer the Features

At this stage, you need to start assigning variables to your data. You need to factor in what will affect your data. Does a heatwave during the summer cause sales to drop? Does the holiday season affect sales in all stores and not just middle-to-high-income neighborhoods? Things like seasonal purchases become variables you need to account for.

Create Your Predictive Models

At some point, you’ll have to come up with predictive models to support your hypotheses. For example, you’ll have to write code to predict sales. You may explore whether an after-Christmas sale increases profits and, if so, by how much. You may find that a certain percentage of sales earns more money than other sales, given the volume and overall profit.

Communicate Your Results

In the real world, all the analysis and technical results you come up with are of little value unless you can explain to your stakeholders what they mean in a comprehensible and compelling way. Data storytelling is a critical and underrated skill that you must develop. To finish your project, you’ll want to create a data visualization or a presentation that explains your results to non-technical folks.

Data Science Projects FAQs

Since you’re here…Are you a future data scientist? Investigate with our free guide to what a data scientist actually does. When you’re ready to build a CV that will make hiring managers melt, join our Data Science Bootcamp that guarantees a job or your tuition back!