Aaron Bateni

Research

My research interest is in developing novel methods and optimization techniques that enable and improve intelligent reasoning within different problem domains. Aiming one day, to make people utilize models like LLMs, ViTs, or even diffusion models like DiTs to solve current tasks, or innovative new tasks efficiently on day-to-day edge devices (e.g. phones and laptops).

Multi-Class Emotion Classification Using Transformers (In progress)

WESAD dataset is a mix of time-series biological signals (such as ECG, EDA, EMG, respiration, etc.) used for emotion classification. While majority of previous works focused on combining all classes to binary classification, we introduce a new method to classify multiple classes. The model Achieves an F1-Score of 0.93, which to the best of our knowledge is the State-Of-The-Art.

Hybrid Spiking Neural Network -- Transformer Video Classification Model

Inspired by hybrid Transformer-CNN models, we design a first-of-its-kind hybrid Transformer-SNN model that is inspired by the Cortical Column structure of the brain. The model is then implemented and tested in time-series classification. This work was a result of my original BSc thesis.

Projects

I have also worked on a variety of personal projects. some of which are featured below. for a full list, I recommend checking out my GitHub page.

Study on Effects of UNet's Variations in Polyp Segmentation

I conducted a study on effects of architectural tweaks in the UNets model architecture. I was particularly interested to see how the height of the UNet affects its performance.

Loss Function Optimization with Particle Swarm Optimization

I designed a PSO method that minimizes a given loss function by deploying searching particles with momentum and density.

Pacman Playing Agent

I developed a Q-Learning agent that learns and plays the Pacman game.

Implementation of Spiking Neural Architectures

Implemented several Neural Architectures as prerequisites to my BSc Project. The list is as follows: (Image Credit: Wikimedia)

1. Neural Models: LIF, ELIF, AELIF
2. Encoding Methods: Poisson, TTFS, Positional
3. Currents: Sinusoidal, Constant, Step Current, Multi Step Current

Extensive hands-on experience with Pytorch.

Comprehensive Analysis of Information Retrieval Algorithms

Comprehensive implementations of several Information Retrieval Systems. The list is as follows:

1. WebIR with boolean queries
2. WebIR with wildcard query (Trie, and Permuterm), and spellchecking (Soundex, and Levenshtein Distance)
3. WebIR with Block-Sorted Based Indexing
4. Ranked WebIR with sorting queries based on relevance (Okapi BM25, Language Model, and TF-IDF)
5. Comprehensive Performance Comparison of IR methods

Review of Ant Colony Methods in SCP

Implementations and comparison of several Ant Colony System methods to solve the Set Covering Problem. (Image credit: Wikimedia)

Simulated Annealing for CSP

I designed a simulated annealing method to solve the classic Cutting Stock Problem.

Genetic Method for N-Queen Problem

I developed an evolutionary genetic algorithm to solve the classic N-Queens problem.

ClassBoom

Effected by the boom of E-Learning during the Covid-19 pandemic, I set out to create an E-Learning platform from scratch. Some features include the ability to upload class recordings, homework, grade students individually and more. I mainly utilized Python and Django.

Corridor Game

Using C++, I developed a client-server program to play the world famous Quoridor Game on lan. (Image credit: Wikimedia)

Teaching & Academic Service

I have also been an active member of our academic community. Some of my main contributions are as follows.

Teaching Assistantship

• TAed for: Design and Analysis of Algorithms, Graph Theory, Logical Circuits & Architecture, Combinatorics, Calculus I, and Basic Programming.
• Held classes for 250+ students and prepared 25+ assignment series.
• Designed real-world problems and aggregated widely used problems from academic references.
• Graded 1000+ student submissions, quizzes and exams.

Elected Member of the Computer Science Students’ Scientific Chapter

• Organized workshops on career development, resume writing, and programming with distinguished speakers.
• Organized orientation for the new students on various topics such as programming languages/disciplines.

Design and source code partially borrowed from Jon Barron's website