Abstract: Two important nonparametric approaches to clustering emerged in the 1970's: clustering by level sets or cluster tree as proposed by Hartigan, and clustering by gradient lines or gradient flow as proposed by Fukunaga and Hosteler. In a recent paper, we argue the thesis that these two approaches are fundamentally the same by showing that the gradient flow provides a way to move along the cluster tree. In making a stronger case, we are confronted with the fact the cluster tree does not define a partition of the entire support of the underlying density, while the gradient flow does. Abstract: Mean shift is a simple interactive procedure that gradually shifts data points towards the mode which denotes the highest density of data points in the region. Mean shift algorithms have been effectively used for data denoising, mode seeking, and finding the number of clusters in a dataset in an automated fashion.

Large pre-trained language models have been shown to store factual knowledge in their parameters, and achieve state-of-the-art results when fine-tuned on downstream NLP tasks. However, their ability to access and precisely manipulate knowledge is still limited, and hence on knowledge-intensive tasks, their performance lags behind task-specific architectures. Additionally, providing provenance for their decisions and updating their world knowledge remain open research problems. Pre-trained models with a differentiable access mechanism to explicit non-parametric memory can overcome this issue, but have so far been only investigated for extractive downstream tasks. This NLP research paper explores a general-purpose fine-tuning recipe for retrieval-augmented generation (RAG) -- models which combine pre-trained parametric and non-parametric memory for language generation.

Oleg Shipitko, Chief Technical Director of Evocargo, an integrated logistics service company using autonomous vehicles speaks with Kate. Oleg talks about the need for automating operational logistics inside enclosed facilities centers and how their autonomous vehicles and other operational services can greatly improve the current way we transport goods within facilities such as ports, warehouses and factories. He has a bachelors and masters degree in autonomous information and control systems (bachelors: 4.96/5.0, Oleg has received numerous awards including: Best paper awarded at 32nd European Conference on Modeling and Simulation (ECMS-2018): Ground Vehicle Localization With Particle Filter Based On Simulated Road Marking Image and Best paper awarded at IV International Conference on Information Technology and Nanotechnology (ITNT-2018): Gaussian filtering for FPGA based image processing with High-Level Synthesis tools.

Petuum's CASL research and engineering team has won this year's OSDI 2021 Best Paper Award. This effort is led by Dr. Aurick Qiao who heads the Composability, Automatic, and Scalable Learning (CASL) research and engineering team at Petuum. Dr. Qiao received the Jay Lepreau Best Paper Award at the 15th USENIX Symposium on Operating Systems Design and Implementation (OSDI) 2021 for the paper he co-authored, Pollux: Co-adaptive Cluster Scheduling for Goodput-Optimized Deep Learning which captures the revolutionary work implemented using one of CASL's key components, AdaptDL. Current live application of Pollux can be implemented via AdaptDL that integrates with PyTorch, Microsoft NNI, and with Ray coming soon. Pollux as implemented by AdaptDL improves scheduling performance in deep learning (DL) clusters by adaptively co-optimizing inter-dependent factors both at the per-job level and at the cluster-wide level.

In this technical talk, Amanda Prorok, Assistant Professor in the Department of Computer Science and Technology at Cambridge University, and a Fellow of Pembroke College, discusses her team's latest research on what, how and when information needs to be shared among agents that aim to solve cooperative tasks. Effective communication is key to successful multi-agent coordination. Yet it is far from obvious what, how and when information needs to be shared among agents that aim to solve cooperative tasks. In this talk, I discuss our recent work on using Graph Neural Networks (GNNs) to solve multi-agent coordination problems. In my first case-study, I show how we use GNNs to find a decentralized solution to the multi-agent path finding problem, which is known to be NP-hard.

A Generative Adversarial Network for AI-Aided Chair Design Researchers present a deep neural network for improving human design of chairs which consists of an image synthesis module and a super-resolution module. They select one of the candidates as a design prototype and create a real-life chair based on it. According to the researcher team, this is the first physical chair created with the help of deep neural networks, which bridges the gap between AI and design. This is the largest NLP model ever trained, with 17 billion parameters. T-NLG has achieved SOTA performance on mainstream NLP tasks.

A panel of judges at the International Conference for High Performance Computing, Networking, Storage and Analysis (SC19) on Thursday awarded a multi-institutional team led by Lawrence Livermore National Laboratory computer scientists with the conference's Best Paper award. The paper, entitled "Massively Parallel Infrastructure for Adaptive Multiscale Simulations: Modeling RAS Initiation Pathway for Cancer," describes the workflow driving a first-of-its-kind multiscale simulation on predictively modeling the dynamics of RAS proteins -- a family of proteins whose mutations are linked to more than 30 percent of all human cancers -- and their interactions with lipids, the organic compounds that help make up cell membranes. Developed as part of the Pilot 2 project in the Joint Design of Advanced Computing for Cancer program, a collaboration between the Department of Energy (DOE) and National Cancer Institute (NCI), the research resulted in a Multiscale Machine-Learned Modeling Infrastructure (MuMMI) that investigators found was scalable to next-generation heterogenous supercomputers such as LLNL's Sierra and Oak Ridge's Summit. Working for more than two years on the pilot project, which is funded by the National Nuclear Security Administration's Advanced Simulation and Computing program, the multidisciplinary team, composed of more than 20 computational scientists, biophysicists, chemists and statisticians from LLNL, Los Alamos National Laboratory, NCI/Frederick National Laboratory for Cancer Research, Oak Ridge National Laboratory (ORNL) and IBM, ran nearly 120,000 simulations on Sierra, using 5.6 million GPU hours of compute time and generating a massive 320 terabytes of data. "I can't begin to describe how happy I am for our team -- it's been a lot of hard work, and to have it recognized at this level is just amazing," said Francesco Di Natale, LLNL computer scientist and the paper's lead author.

There are only a handful of machine learning conferences in the world that attract the top brains in this field. One such conference, which I am an avid follower of, is the International Conference on Machine Learning (ICML). Folks from top machine learning research companies, like Google AI, Facebook, Uber, etc. come together and present their latest research. It's a conference any data scientist would not want to miss. ICML 2019, held last week in Southern California, USA, saw records tumble in astounding fashion.

DeepMind's New AI Tracks Serengeti Herds from Images Alone DeepMind, the U.K.-based AI research subsidiary acquired by Alphabet in 2014 for $500 million, today detailed ecological research its science team is conducting to develop AI systems that'll help study the behavior of animal species in Tanzania's Serengeti National Park. They extend the popular BERT architecture to a multi-modal two-stream model, processing both visual and textual inputs in separate streams that interact through co-attentional transformer layers.

The two-day 21st Malaysian Finance Association Conference 2019 held at Sunway University was one of the most important finance research conferences in Malaysia. The conference provided an open platform for scholars and practitioners in the field to meet and share their current research as well as to exchange ideas and information on new developments in the areas of finance and economics. The conference was officially launched by Professor Catherine Ho Soke Fun, Malaysian Finance Association President; Professor Graeme Wilkinson, Sunway University Vice-Chancellor, and Chong Chang Choong, Sunway Group Chief Financial Officer. Chong in his opening address said, "We live in a digital age where internet and technology have become a big part of our lives. The rapid advancement of technology is transforming the workplace and the world, including the financial sector. Fintech or financial technology, is now no longer a peculiar term. The ever-evolving Fintech services are beginning to be adopted by the tech-savvy and sophisticated consumers of today."