Personal
A Verification Framework for Component-Based Modeling and Simulation Putting the pieces together
In this thesis a comprehensive verification framework is proposed to contend with some important issues in composability verification and a verification process is suggested to verify composability of different kinds of systems models, such as reactive, real-time and probabilistic systems. With an assumption that all these systems are concurrent in nature in which different composed components interact with each other simultaneously, the requirements for the extensive techniques for the structural and behavioral analysis becomes increasingly challenging. The proposed verification framework provides methods, techniques and tool support for verifying composability at its different levels. These levels are defined as foundations of consistent model composability. Each level is discussed in detail and an approach is presented to verify composability at that level. In particular we focus on the Dynamic-Semantic Composability level due to its significance in the overall composability correctness and also due to the level of difficulty it poses in the process. In order to verify composability at this level we investigate the application of three different approaches namely (i) Petri Nets based Algebraic Analysis (ii) Colored Petri Nets (CPN) based State-space Analysis and (iii) Communicating Sequential Processes based Model Checking. All three approaches attack the problem of verifying dynamic-semantic composability in different ways however they all share the same aim i.e., to confirm the correctness of a composed model with respect to its requirement specifications.
Hierarchical Federated Learning with Quantization: Convergence Analysis and System Design
Liu, Lumin, Zhang, Jun, Song, Shenghui, Letaief, Khaled B.
Federated learning (FL) is a powerful distributed machine learning framework where a server aggregates models trained by different clients without accessing their private data. Hierarchical FL, with a client-edge-cloud aggregation hierarchy, can effectively leverage both the cloud server's access to many clients' data and the edge servers' closeness to the clients to achieve a high communication efficiency. Neural network quantization can further reduce the communication overhead during model uploading. To fully exploit the advantages of hierarchical FL, an accurate convergence analysis with respect to the key system parameters is needed. Unfortunately, existing analysis is loose and does not consider model quantization. In this paper, we derive a tighter convergence bound for hierarchical FL with quantization. The convergence result leads to practical guidelines for important design problems such as the client-edge aggregation and edge-client association strategies. Based on the obtained analytical results, we optimize the two aggregation intervals and show that the client-edge aggregation interval should slowly decay while the edge-cloud aggregation interval needs to adapt to the ratio of the client-edge and edge-cloud propagation delay. Simulation results shall verify the design guidelines and demonstrate the effectiveness of the proposed aggregation strategy.
SEVENPOINTONE Awarded Innovation Honoree at CES 2023 - Startup World Tech
On November 16 2022, SEVENPOINTONE announced that it would receive the title of Innovation Awards Honoree at CES 2023. The award was presented to them for "AlzWIN", an AI-powered service for the early detection of dementia and other cognitive impairments, that is simple to apply and quick to diagnose. CES is the one of the largest tech events in the world, where innovators come together to showcase their services, many of which have traveled from across the world to attend. In Las Vegas USA, they can meet with prospective business partners, exchange ideas and enjoy the presentations. AlzWIN recognizes the fluency of the users speech, as well as the semantic memory within two minutes of the analysis.
Touching - or grotesque? The AI gadget that lets you 'talk' to loved ones beyond the grave
It sounds like something out of a dystopian sci-fi film: virtual-reality humans that mimic the voices and facial expressions of dead people, so loved ones can speak to them from beyond the grave. But now a Korean artificial intelligence firm has made that'ghoulish' vision a reality, insisting the controversial service will help friends and family cope with grief. The technology uses machine learning to process images, audio recordings and video footage of recently deceased people to create a virtual version that can interact with the living, as if on a video call. Called re;memory, the software can even answer questions from the grieving, and share memories from the past. It is expected to cost between ยฃ10,000 and ยฃ20,000 to create a virtual person, then ยฃ1,000 each time a loved one wants a conversation with it. Business development manager Joseph Murphy acknowledged that the service was controversial.
Artificial Intelligence: What Is Real?
What do you think of images that attempt to mirror reality? I was talking with a woman recently who retired ten years ago after a career as the head of technology in her company. She's now 77 and realizes how much the field has changed over the past decade and she now wants to dig down to learn about how artificial intelligence is transforming the world.
RLAS-BIABC: A Reinforcement Learning-Based Answer Selection Using the BERT Model Boosted by an Improved ABC Algorithm
Gharagozlou, Hamid, Mohammadzadeh, Javad, Bastanfard, Azam, Ghidary, Saeed Shiry
Answer selection (AS) is a critical subtask of the open-domain question answering (QA) problem. The present paper proposes a method called RLAS-BIABC for AS, which is established on attention mechanism-based long short-term memory (LSTM) and the bidirectional encoder representations from transformers (BERT) word embedding, enriched by an improved artificial bee colony (ABC) algorithm for pretraining and a reinforcement learning-based algorithm for training backpropagation (BP) algorithm. BERT can be comprised in downstream work and fine-tuned as a united task-specific architecture, and the pretrained BERT model can grab different linguistic effects. Existing algorithms typically train the AS model with positive-negative pairs for a two-class classifier. A positive pair contains a question and a genuine answer, while a negative one includes a question and a fake answer. The output should be one for positive and zero for negative pairs. Typically, negative pairs are more than positive, leading to an imbalanced classification that drastically reduces system performance. To deal with it, we define classification as a sequential decision-making process in which the agent takes a sample at each step and classifies it. For each classification operation, the agent receives a reward, in which the prize of the majority class is less than the reward of the minority class. Ultimately, the agent finds the optimal value for the policy weights. We initialize the policy weights with the improved ABC algorithm. The initial value technique can prevent problems such as getting stuck in the local optimum. Although ABC serves well in most tasks, there is still a weakness in the ABC algorithm that disregards the fitness of related pairs of individuals in discovering a neighboring food source position.
Ukraine war: What does facial recognition software make of Putin's backdrop crowd?
The woman at the New Year address has been named in Russian media as Anna Sergeevna Sidorenko, a captain and military doctor. Comparing her face at the event with an image taken from a video interview posted online by the Russian Isvestia newspaper gave a 99.5% match. Her name also appears on a members list of a Russian military regiment published by the Ukrainian intelligence services.
WAYCEN wins CES innovation awards among medical AI companies - Coleda Pvt Ltd
AI Medtech company WAYCEN (CEO Kyungnam Kim) has won four CES innovation awards among medical AI companies for the first time ahead of "CES 2023", the world's largest electronics trade show, which will be held in Las Vegas, USA, in next January. According to CES 2023 awards results announced by the Consumer Technology Association (CTA) on February 16, WAYCEN has won four innovation awards for WAYMED Cough, WAYMED Endo PRO and WAYMED EBUS. With two awards each in the areas of digital healthcare and software & mobile apps, it demonstrated the technological leadership of K-medtech. Winning these awards means that WAYCEN has been recognized worldwide for its technical expertise in real-time image analysis technology and medical big data analysis technology. Winning two awards exclusively for the products using real-time image analysis technology is intended to highlight innovations in WAYCEN's medical AI solution range.
$3,300 self-driving stroller latest high-tech product for parents
White Castle Vice President Jamie Richardson shows how'Flippy' the robot flips burgers and drops fresh fries at 100 locations nationwide. An artificially intelligent stroller with a hefty price tag is the latest innovation announced to aid parents. Canadian startup Glรผxkind Technologies announced that it had been named a CES 2023 Innovation Awards honoree for its smart stroller "Ella." The product, with a purchase price starting at $3,300, includes environmental danger alerts, intelligent braking and 360-degree monitoring. In a press release, Glรผxkind noted the stroller's ability to push and brake regardless of terrain and even when fully loaded with groceries and toys.
"Machine Learning and healthcare are well-suited to one another" - Elets eHealth
The entire value chain of the healthcare industry is undergoing a huge transformation on account of technology adoption in reducing the cost of healthcare, and accessibility of healthcare facilities, shares Gautam Khanna, Chief Executive Officer, P D Hinduja Hospital & Medical Research Centre with Kaanchi Chawla of Elets News Network (ENN). What position do you think Indian healthcare start-ups are in right now, and how are they doing? India's start-up ecosystem is thriving. The country is home to over 100 Unicorns. According to the NASSCOM Startup Ecosystem Report 2015, India serves as the fastest growing startup-base worldwide and six to eight per cent of the recent B2C startups in India belong to the healthtech sector.