AI, and machine learning (ML) in particular, are very important job skills.
The most progressive IT companies — from small startups to unicorns — need people who have AI and ML skills. These people work on fascinating projects (look at this AI jobs list to have an idea) and enjoy nice compensation and professional growth opportunities.
If you’re looking to join these highly skilled professionals, creating test projects is a good idea for practice. Test projects can be comparably easy but give you valuable job skills, so listing them on your resume would be a huge plus.
Need some ideas for test projects?
These five easy AI and ML learning projects should be excellent additions.
1. Fake News Detection Model
One in two American adults admits to sharing fake news online.
Just think about that—
Almost half of the population at some point has found super persuasive fake news that was believable enough to share with their family and friends. The epidemic of fake news is only getting stronger, and we need effective ways to minimize its impact.
That’s why creating a fake news detection model would be a good way to practice skills and contribute to the effort.
You can create such a model in two ways.
The first is to use Natural Language Processing (NLP) to detect words and phrases that strongly suggest that a piece of news is fake.
The second one is to use the text classification method. Accordingly, you would make two datasets of both real and fake news to create an ML model for differentiating credibility based on the learned patterns.
Having this project on your resume would be a good addition. If you’d like to apply for AI jobs that involve text analysis, a fake news detection experience would be a plus.
2. Stock Price Prediction
Have you heard about FinTech, the application of AI in the financial sector?
The FinTech startup scene is booming right now, but the talent shortage is becoming a real concern. In the UK alone, 24% of FinTech companies face a shortage of skills—and the situation is similar in other countries.
So, your timing couldn’t be better if you’d like to work in FinTech. One way in which you can demonstrate valuable skills is to build a model for stock price prediction.
This model is good because it:
- Shows your expertise in working with complex and fast-changing data sources
- Confirms your predictive analytics skills
- Showcases your ML and AI expertise
To build a stock price prediction model—
- Get data — get historical price data for a single stock of your choice for several years (Yahoo Finance and MarketWatch are fantastic resources that’ll allow you to get the data within 20 seconds)
- Run a regression function — use a neural network to run a regression function to split numerical values in price data
- Build an ML model — install TensorFlow Artificial Neural Network (TFANN), an open-source library for data flow programming, and import dependencies, and add the historical price data.
- Train the model to predict prices — try different numbers of days, price change trends, and other indicators to train the model to understand the real price changes
Eventually, you’ll create a model that will be able to generate predictions on stock prices based on real historical data.
3. Social Media Sentiment Analyzer
Brand reputation is an important thing for all companies. That’s why they invest in social media sentiment analysis apps — tools that analyze social media posts to learn about what people are saying about their brands.
That’s why they hire data scientists, analytics, and other specialists with AI skills to help them manage reputation.
You can be one of them!
If you search for internships or entry-level jobs with the “Sentiment analysis” keyword on AI Jobs, you’ll find positions at such companies as Coursera, HelloFresh, and Kraken.
You can create an ML system to analyze the sentiment behind social media posts, too.
Basics steps are:
- Collect data from a chosen social media network (Zapier, Content Grabber, and Scrape Storm are good tools for that).
- “Clean” the data to remove special characters, irrelevant data, web links, incorrectly spelled words, etc.
- Use Jupyter Notebook or Google Colab for interactive analysis of the datasets you created.
- Assign special values to posts to differentiate between “positive,” “negative,” and “neutral” ones.
There are also pre-made social media sentiment analysis tools online you can check to understand how their own algorithms work. If you add this project to your resume, that’ll be a great step to optimize it for a data scientist or analyst career.
4. Face Recognition Algorithm
Face recognition is a highly valuable feature of many applications today, so having the skills to create a similar one is a big plus.
You can learn face recognition with Python — this programming language has a special library designed for this goal. It has a ‘face_locations’ command you can use to try and detect the faces of individuals on pictures.
The basics of using face recognition in Python:
- Install libraries (OpenCV and NumPy are good options) — they are required
- Configure a camera and connect it to your computer (this involves camera training with a cascade function)
- Collect photos of people for the experiment and ID each person in them (try one person per photo)
- Train the ML model to recognize images. This means adding photos to a chosen library and adding test codes.
The end result should be an ML face detection and recognition model in Python. As you build your skills, consider adding photos of multiple individuals in one photo to challenge the model.
5. Build a QR Code Scanner
A QR scanner is a simple but highly useful application used in marketing, education, healthcare, and other industries. That’s why there are QR readers and scanners made for every platform, and the technology keeps evolving.
Similar to the face recognition model, you’ll need Pillow, Pyzbar, and OpenCV libraries to build a QR code scanner.
Steps to build a QR code reader in Python:
- Install the three libraries (Pillow, Pyzbar, and OpenCV)
- Write the decoding function to teach the model to understand and interpret information on QR codes
- Write the main function for turning on the camera and using the decoding function to read QR codes
The end result should be a model able to use a camera to read and understand QR codes. Adding this simple yet useful experience to your resume will be another plus for your candidacy that recruiters will consider.
AI and Machine Learning Projects for Students: Summary
There are so many fascinating projects you can do even with basic AI and ML skills. These five options, for example, are good ones to impress recruiters and demonstrate your practical relevant knowledge.
Because the end result of each project is something useful and relevant in today’s world, having at least some on a resume would be helpful to get callbacks and interviews.
WRITTEN BY Alison Lee
Alison Lee is an experienced educational technology writer. For the past five years, she has been sharing her knowledge on writing, ultimately creating a database of academic writing examples: https://subjecto.com/essay-samples/. When she’s not writing, she likes to participate in online EdTech conferences and meetups.
PHOTO BY Jeswin Thomas