International Speakers

Ali Kashif Bashir (University of the Faroe Islands, Denmark)
Internet of Things: Prospects, Challenges and Use Cases (Cryptocurrency)

Abstract: Internet of things (IoT) has revolutionized the way we connect and communicate with devices. IoT capabilities are considered as ‘game changing’ when applied with the concepts of big data, cloud, block chain, etc. In this talk, we will discuss the impact of IoT on our daily lives in terms of healthcare, surveillance, etc. We will also discuss its prospects and challenges in relation to other technologies such as clouds, data centers, etc. In order to explain its impact, we will consider uses cases like smart homes, healthcare and cryptocurrencies. I will take the leverage to introduce our initiatives in implementing IoT in healthcare and cryptocurrencies. In cryptocurrency, I will introduce our new start-up named Bandz Network: A P2P Bandwidth Economy Extranet.

Humaira Nisar (Universiti Tunku Abdul Rahman, Malaysia)
Analysis of Effect of Music on Human Brain using Electroencephalography

Abstract: The aim of this study is to investigate the effect of music stimuli on human brain using EEG. The study comprises a short term and a long term experiment. Two types of music stimuli; favorite music (preferred music of the participating subjects) and relaxing music (composed of alpha binaural beats) are used. Assessment of the soothing effects of music on human brain is done by analyzing different features; absolute power in the alpha band, approximate entropy, sample entropy and frontal asymmetry using EEG recordings. Classification of different signals is done by using deep learning methods. The results from the study indicate that relaxing music has more relaxing effects on the human brain than the favorite music.

Nasir M. Rajpoot (University of Warwick, UK)
Cancer Histology Footprint Analytics

Abstract: The human brain is fantastic at recognising people and objects and building an understanding of the natural world around us. However, the visual cortex is not ideal at objectively measuring what we see and complex spatial patterns hidden in plain sight cannot sometimes be deciphered by the naked eye. Computational Pathology is an emerging discipline concerned with the study of computer algorithms for understanding disease from the analysis of digitised histology images. I will show some snippets of computational pathology research in my group to demonstrate the value of analytics of information-rich, high-resolution whole-slide images (WSIs, the so-called Big Cancer Image Data) for cancer diagnosis and prognosis. I will show examples of how histological motifs extracted from digital pathology image data are likely to lead to patient stratification for precision medicine. I will then discuss some of the main challenges in digital pathology and opportunities for exploring new unchartered territories.

Prof. Rashid Mehmood (King Abdul Aziz University Jeddah, KSA)
Smart Cities Technologies, Infrastructure and Applications

Abstract: Smart cities provide next generation engineering approaches for urbanization, having evolved from knowledge-based economy, digital economy and intelligent economy. Smart cities aim to exploit the intellectual and social capital as its core ingredient for urbanization, in addition to the physical and ICT infrastructure. Smart cities are driven by several interdependent trends. These include a pressing need for environmental sustainability, and peoples’ increasing demands for personalization, mobility and higher quality of life. Technological developments such as miniaturisation of devices, internet of things (IoT), big data, computational and artificial intelligence, and decreasing costs of computational entities have also accelerated the smart cities developments. Smart cities encompass all aspects of modern day life, transportation, healthcare, entertainment, work, businesses, social interactions, governance, etc. It is therefore necessary to engineer smart cities as complex systems of systems supported by a converged ubiquitous infrastructure. A key challenge in the realisation of smart cities is to create an ecosystem of digital infrastructures that are able to work together and enable dynamic real-time interactions between various smart city subsystems. In this talk I would trace the evolution of smart cities with the aim to explore the state of the art of infrastructure underlying smart cities. The technologies focus would be on IoT, big data, AI and high performance computing.

M. Omair Shafiq (Carleton University, Ottawa, ON, Canada)
Role of Data Modeling and Analytics for Emerging Big Data Applications

Abstract: Monitoring and management of large-scale applications have always been complex tasks, especially because execution data is logged in software applications in an unstructured and human-readable manner. This makes it limited, requires manual interpretation and makes the process of monitoring and management to be slow, cumbersome and hard. This talk will present an overview of our proposed solution of data modeling in correlation with analytical techniques in order to improve monitoring and management process for emerging large-scale, complex and data-intensive software applications (i.e., big data applications).

National Speakers

Hammad Naveed (NUCES-FAST, Islamabad, Pakistan)
The era of computation in Biology

Abstract: As more and more biological data is accumulated due to the advancement in experimental technologies, computing is starting to play a major role in biological studies. I will briefly introduce the work being done in the Computational Biology Research Lab (CBRL) at FAST-NU Islamabad. This will cover i) 3D structure modeling of membrane proteins ii) using machine/deep learning to solve biology questions and iii) social data mining to solve biology problems. I will give a detailed talk on structure based identification of new targets for existing drugs. Current cancer drugs have significant side-effects, many of which might be due to the off-targets of these drugs. I will present a novel data science based approach for large-scale discovery of new targets for existing drugs. For a given drug, I construct the 3D signature of drug binding sites that captures its promiscuous structural features. I am able to predict the interaction of six cancer drugs with their known targets with a sensitivity of 50% and specificity of 56%. Computational predictions were validated by in vitro binding experiments. This method is broadly applicable for the prediction of protein-drug interactions with novel applications to biological research and drug development.

Muhammad Muzammal (Bahria University, Islamabad, Pakistan)
Decentralized Trustless Applications: Challenges and Opportunities

Abstract: Blockchain technology is increasingly becoming prevalent in application domains that require trustless exchanges in decentralized environments. As centralized business applications are based on the notion of trust, they are prone to occasional security and privacy breaches. Further, the integrity of the centralized systems is often undermined by the adversaries. However, the integrity and immutability of the blockchain along with data security and privacy creates a host of new applications. For example, in a supply chain environment, a product can be tracked in a trustless environment, finances can be handled without intermediaries, and money transfers are quick; in healthcare, counterfeit medicine can be tracked and identified; blockchain can revolutionize smart grid by enabling trustless energy exchanges; in artificial intelligence, blockchain technology can be used to design accountable agents; user-controlled data access has the potential to radically transform the social networking as we know it today; data generated by IoT devices and the privacy preservation in blockchain systems opens venues for privacy-preserving big data analytics;  and many more. However, blockchain systems lack in query processing and throughput. Therefore, novel theories are needed which enable fast query processing and high throughput for blockchain systems. Similarly, the applicability of the blockchain technology in the aforementioned domains is although exciting but is not obvious and requires interesting formulations which can address the limitations in the current blockchain and enterprise systems, simultaneously. This talk is about visions for blockchain systems in interesting application domains and gives useful insights into the blockchain technology and applications.

Dr. Ahmad R. Shahid (CIIT, Islamabad, Pakistan)
Explainable AI (XAI)

Abstract: AI is notorious for being opaque. It is not easy to explain the decisions taken by the AI algorithms. Even their designers cannot explain it. The requirement for explainability has increased manifold with the advent of deep learning, where many layers of neurons say something together and the only information available is their weights which are not intuitively clear. However, there is no free lunch. Normally explainable AI comes at the cost of performance. In future, the aim is to build systems that can explain what they have done, why they have done, and take feedback from the user to improve further.

Sibt ul Hussain (NUCES-FAST, Islamabad, Pakistan)
Deep Learning and its Practical Applications

Abstract: In this talk i will be introducing the reveal.ai and contributions it is making in the domains of deep learning and reinforcement learning for building real life intelligent systems ranging from intelligent EHRs to autonomous vehicles. Specifically, i will be introducing our state-of-the-art deep learning and reinforcement learning systems for: (i) predicting procedures and drugs from the patient historical data; (ii) learning and modeling drivers behaviours for driving autonomous vehicles; (iii) generating super-resolution and deblurred images from low-resolution ones; (iv) Urdu text recognition and understanding; (v) learning homography matrices to perform image deformation and dewarping, etc.