Recently, Wolfram Research organized a virtual winter school in India. Despite many issues caused by the omicron surge that impacted some of our mentors and students, we successfully completed a three-week remote computational school where each student completed a project under the supervision of a mentor.

How It Worked

This school was based on the experiences and resources we have developed at Wolfram under the umbrella of our summer programs. We provided a hybrid system of recorded lectures and live interactions with a strong team of mentors and teaching assistants (TAs).

The first phase of the India school consisted of preparation and training that we conducted asynchronously. Recordings of lecture series by top developers at Wolfram Research from previous educational programs were carefully selected and shared with students.

To keep students engaged and help them with problem-solving skills using the Wolfram Language, we assigned them a series of challenges across many different fields (e.g. from geography and numerical calculations to machine learning). Students were highly engaged with our team of TAs, who were students from our previous educational programs. Having these TAs provide support and effective engagement was among the key factors for our success.

After assessing our students and helping them with the first phase, we paired each student with a mentor. The pairings were based on students’ interests and potential projects they would like to do. Our mentors provided the project details; in fact, each mentor shared a series of potential computational projects with students, and students had to select a few they’d be interested in exploring.

The next two weeks were comprised of active interaction and engagement between mentors and students, together with the help of our TA team. Toward the end of the third week, students wrote their projects in Wolfram Notebooks in the style of a computational essay and submitted them to Wolfram Community. You can also find the collection of notebooks in the Wolfram Notebook Archive.

A Sample of Student Projects

Let’s take a look at some of the standout projects students completed with guidance from mentors and TAs.

Flow Visualizations with Navier–Stokes Equations »

Abhishek Joshi

Abhishek Joshi is a master’s student at the LNM Institute of Information Technology and is currently working on topological insulators.

In his project, Joshi did benchmark computations testing 2D flow over a cylinder in a flat channel in transient and steady cases. Navier–Stokes equations can be used to determine the velocity vector field that applies to a fluid, given some initial conditions. He did a visual analysis for the solutions of the equations and found that the benchmark computations for laminar flow were agreeable with the theoretical predictions.

Computing Nuclear Properties in the FCC Model »

Adiba Shaikh

Adiba Shaikh is a doctoral student at the Indian Institute of Technology Bombay and works with various high energy physics.

Shaikh’s project used a face-centered cubic (FCC) structure for the arrangement of protons and neutrons within the nucleus. She implemented the FCC model for nuclear structure and computed some of the basic nuclear properties, such as number density, quantum numbers, alpha particle structure and the stability of the nuclei having magic numbers of nucleons. Comparing the binding energy for various nuclei indicated their enhanced stability due to the presence of magic numbers of protons and neutrons.

Interactive and Dynamic Financial App »

Aryan Barapatre

Aryan Barapatre is currently a sophomore at Birla Institute of Technology and Science, Pilani and is pursuing degrees in mathematics and computer science.

Barapatre developed an interactive interface that provides real-time information about stocks in an intuitive format for users. After successfully extracting real-time data from websites like Yahoo! Finance, he was then able to plot the dynamic data into charts and create an interactive interface to display the stock data. Other test cases where this method can be useful include analyzing cryptocurrency and Tour de France competition data in real time.

Analyzing Unemployment in India during COVID-19 Pandemic »

Ashika Deb

Ashika Deb is pursuing a bachelor’s degree in mathematics with a minor in computer science at Miranda House, University of Delhi.

In her project, Deb studied unemployment data in India to examine relationships between various factors that may have affected unemployment during the COVID-19 pandemic. Unemployment gives birth to multiple social and economic problems, and the pandemic exposed and exacerbated existing inequities in the labor market. After examining available data from official sources, an exploration analysis using the Wolfram Language drew out meaningful statistics.

Radiation Leak Detector »

Deepak Yadav

Deepak Yadav is a master’s student in physics at the University of Mumbai.

In this project, Yadav explored data from nuclear reactor detectors that can be lowered through the axis of the reactor container and record radiation levels at multiple points. After designing and training a classifier, synthetically generated data used Monte Carlo simulations based on a basic geometry, including a fuel pin, helium surrounding, Zircaloy cladding and water. Future work includes training a classifier to identify a leak by looking at the data and also predicting the approximate leak position.

Emotion Classification from Face Datasets »

Diya Elizabeth

Diya Elizabeth is a student at the Indian Institute of Science Education and Research Thiruvananthapuram and is pursuing an integrated master’s degree in mathematics with a minor in data science.

In her project, Elizabeth used deep learning to train an emotion recognition neural network using the Facial Expression Recognition 2013 (FER-2013) Dataset available in the Wolfram Data Repository. Seven classes of emotions were considered in the dataset—happy, sad, angry, surprise, disgust, fear and neutral—and different methods were tried out to tackle class imbalances. This model can be applied to teaching autistic children, young children in general and people blind to facial expressions. It can also be used to train robots to interact more efficiently with humans.

Predicting the Electron Invariant Mass from Collision Data »

Shivam Sawarn

Shivam Sawarn earned his bachelor’s degree in physics from Jamia Millia Islamia and is pursuing a master’s degree from the University of Delhi.

In this project, Sawarn explored various features of the dielectron collision using rigorous data analysis tools. The dataset, consisting of 100,000 dielectron events in the invariant mass range 2–110 GeV, was also used to train the neural network. The neural network was used on the test dataset to predict the invariant mass of the electron instead of just taking the mean of the invariant mass dataset.

Cryptocurrency Price Prediction with Recurrent Neural Networks »

Shivay Nagpal

Shivay Nagpal is an undergraduate student studying mathematics and computer science at St. Stephen’s College, University of Delhi.

Nagpal’s project explored a long short-term memory (LSTM) recurrent neural network to make short-term cryptocurrency predictions using the Wolfram Language. The project dataset included market values for three popular cryptocurrencies on the market—Bitcoin (BTC), Ethereum (ETH) and Litecoin (LTC)—and a resource function from the Wolfram Function Repository obtained the data.

Smart Backgrounds in Simulation Data Using Neural Networks »

Utkarsh Patel

Utkarsh Patel is currently enrolled in a PhD program in the high energy physics department at the Indian Institute of Technology Bhilai.

The goal of Patel’s classification project was to identify generated events from the Belle II simulation, a particle physics experiment designed to study the properties of B mesons and other particles. The events first pass a selection before the expensive simulation and reconstruction to visualize the various parameters associated with each particle in graphical form. The event datasets along with other information are available as numpy arrays in the GitHub repository.

More Student Projects

Check out additional student projects from the Wolfram India School 2022 at Wolfram Community:

• “Identifying Galaxies, Quasars and Stars with Machine Learning”
• “An Interface for Julia Language to Connect with the Wolfram Language”
• “Neural Network to Study Cosmic Air Showers”
• “Traffic Sign Recognition Using CNN”
• “Dynamics of Public and Private Opinions”
• “PINN: Physics Informed Neural Network to Predict Motion of 1D SHM”
• “Calling the Wolfram Language from R”
• “Exploring the Wolfram Quantum Computation Framework”