Machine learning and deep learning algorithms have led to good human activity recognition and behavior recognition results, such as stress, depression, and anxiety. We have examined the effects of wearable and survey data on students’ grades, stress, anxiety, and sleep. In this project, we want to add different modalities in the scope of the currently employed dataset, Nethealth, which will be communication-based information.
In this project, a blockchain based Token Rental system will be developed using Solidity smart contracts. The users will be able to (i) post token rental intervals (ii) search/explore posted token intervals and (iii) bid for token rental intervals. A web interface will also be developed for the rental system. The system will be deployed and tested on test chains of Avalanche, Polygon Matic , Binance Smart Chain and Hedera which support Solidity smart contracts.
We have been working mostly on behavioral analysis from wearables using machine learning and deep learning techniques. So far, we have applied our algorithms by not considering the personal effects on the results. This project will focus on personalizing the gathered results by applying transfer learning techniques.
In this project, tools will be developed that will collect and analyze transactions from the widely used Decentralized Exchange (DEX) smart contracts on major blockchains (such as Ethereum, Polygon Matic, Hedera, Avalanche and Binance Smart Chain). The project will develop (i) python based and (ii) web based tools that retrieve and extract transaction data and analyze it by forming transaction graphs.
Graph partitioning algorithms partition a graph into equal sized subgraphs (in terms of nodes) in such a way that the the number of edges cut is minimized. In the more general problem, edges can also have weights. In this project, you will develop a parallel software for spectral graph partitioning.
The aim of this project is to build a desktop tool to be used to analyse and visualize Turkish maqam music pieces. This is an engineering project, it does not involve any research. The skills required are python and docker.
This project aims to develop a tool that analyzes the functionality and dependencies of different software components in a vehicle and recommends the optimal boundaries for dividing the system into microservices that follow the publisher-subscriber pattern. The goal of this project is to improve the scalability, flexibility, and maintainability of software-defined vehicles by breaking down monolithic systems into smaller, modular, and independently deployable services that can communicate using the publisher-subscriber architecture.
Neural Architecture Search (NAS) is a method fpr automating the process of designing neural network architectures by using algorithms such as evolutionary algorithms or reinforcement learning to search through a predefined space of architectures. NAS is used by a couple of open-source AutoML tools such as AutoKeras, AutoPytorch and AutoGluon in order to find a DNN model with best performance for a given task.