Skip to main content

Projects

CHIST-ERA PROJECT, EXPECTATIONPersonalized Explainable AI for decentralized agents with heterogeneous knowledge  (Start Date: 1/07/2021)

Explainable AI (XAI) has recently emerged proposing a set of techniques attempting to explain machine learning (ML) models. The recipients (explainee) are intended to be humans or other intelligent virtual entities. Transparency, trust, and debuging are the underlying features calling for XAI. However, in real-world settings, systems are distributed, data are heterogeneous, the “system” knowledge is bounded, and privacy concerns are subject to variable constraints. Current XAI approaches cannot cope with such requirements. Therefore, there is a need for personalized explainable artificial intelligence. We plan to develop models and mechanisms to reconcile sub-symbolic, symbolic, and semantic representations leveraging on the agent-based paradigm. In particular, the proposed approach combines inter-agent, intra-agent, and human-agent interactions to benefit from both the specialization of ML agents and the establishment of agent collaboration mechanisms, which will integrate heterogeneous knowledge/explanations extracted from efficient black-box AI agents. The project includes the validation of the personalization and heterogeneous knowledge integration approach through a prototype application in the domain of food and nutrition monitoring and recommendation, including the evaluation of agent-human explainability, and the performance of the employed techniques in a collaborative AI environment.

TÜBİTAK Kariyer Projesi, "İnsan-Robot Müzakereleri İçin Otonom Etmen Tasarımı, Geliştirilmesi Ve Deneysel Değerlendirilmesi", Başlangıç Tarihi: 28/11/2018

The ultimate goal of this project is to develop autonomous negotiating agents taking human factors into account for a social robot that needs to negotiate with people, to provide integration of the software agent with the robot, and to demonstrate the effectiveness of the developed agents through human-robot negotiation experiments. In order to achieve this aim, it is aimed at (a) designing a negotiation protocol that determines rules of encounter between human negotiator and robot in negotiation, (b) developing negotiating agents that are sensitive to the feelings of their opponent and also developing negotiating agents that do not consider their opponent’s emotion during the negotiation, and (c) evaluating their performance in human-robot negotiation experiments. In addition, it is planned to investigate the effect of physical embodiment in the human-agent negotiations and the use of gestures by robot on negotiation outcome.

M2M Grid Project: Smart M2M Grids – M2M Internet for dynamic M2M Information Business ecosystem - ITEA 2 European Project (Nov 2014- May 2018)

The Smart M2M grids project is focused on creating enablers for a dynamic cyber-physical information business ecosystem connecting the physical world with the business processes of companies in real-time. The first goal is to connect physical world sensors, actuators and various embedded devices and machines (physical M2M objects) with IT systems automatically/ semi-automatically by applying and extending horizontal open standards based M2M infrastructures for communication and services. The second goal is to enable information management for embedded and distributed application for smart interaction with physical M2M objects and IT back-office systems. The third goal is enabling smart information exchange between selected business cases related to energy, buildings, transportation and consumer M2M products and services to make the future world smart, smooth and secure for consumers/prosumers. The resulting system is aimed at boosting transfer towards a more sustainable society and a novel real-time service economy within selected industrial business cases.

 

Thesis

  1. Yousef  Razeghi, "Agent Based Negotiation For Incentive Driven Privacy Preserving Information Sharing", M.Sc. Thesis, Özyeğin University, 2019.
  2. Maryam Saleki, "Risk-Calibrated Evidential Classifiers", M.Sc. Thesis, Özyeğin University, 2020.
  3. Berkay Turkgeldi, "Familiarity Effect on Human-Agent Negotiations", M.Sc. Thesis, Özyeğin University, 2020.
  4. Ozan Yavuz, "An Agent-based Negotiation Approach for Supply Chains", M.Sc. Thesis, Özyeğin University, 2020.