Data Science Insider: January 8th, 2021

In this week’s Super Data Science newsletter: AI Predictions for 2021. Deepfake Queen Causes Controversy. AI Chipmaker Graphcore Raises $222m as it Takes on Nvidia. Python Set to be Named 'Programming Language of the Year'. Robots Celebrate 2021. Cheers, - The SuperDataScience Team P.S. Have friends and colleagues who could benefit from these weekly updates? Send them to [this link]( to subscribe to the Data Science Insider. --------------------------------------------------------------- [ML Speeds Up Tech Design Process by a Year]( brief: A research team at Sandia National Laboratories has successfully used ML to complete cumbersome materials science calculations more than 40,000 times faster than normal. Their results could herald a dramatic acceleration in the creation of new technologies for optics, aerospace, energy storage and potentially medicine while simultaneously saving laboratories money on computing costs. Sandia researchers used ML to accelerate a computer simulation that predicts how changing a design or fabrication process will affect a material. A project might require thousands of simulations, which can take years to run. The team clocked a single, unaided simulation on a high-performance computing cluster with 128 processing cores (a typical home computer has 2-6) at 12 minutes. With ML, the same simulation took 60 milliseconds using only 36 cores–equivalent to 42,000 times faster on equal computers. This means researchers can now learn in under 15 minutes what would normally take a year. Why this is important: The research could prove widely useful because the simulation they accelerated describes a common event — the change, or evolution, of a material’s microscopic building blocks over time. ML previously has been used to shortcut simulations that calculate how interactions between atoms and molecules change over time. The published results, however, demonstrate the first use of ML to accelerate simulations of materials at relatively large, microscopic scales, which the Sandia team expects will be of greater practical value to scientists and engineers. [Click here to find out!]( [AI Accelerated by Light]( brief: Current computing relies on electrical current passed through circuitry on ever-smaller chips, but in recent years this technology has been bumping up against its physical limits. To facilitate the next generation of computation-hungry technology such as AI and autonomous vehicles, researchers have been searching for new methods to process and store data that circumvent those limits, and photonic processors are the obvious candidate. Featuring scientists from universities across the world, the team developed a new approach and processor architecture. The photonic prototype essentially combines processing and data storage functionalities onto a single chip – so-called in-memory processing, but using light. The team combined integrated photonic devices with phase-change materials (PCMs) to deliver super-fast, energy-efficient matrix-vector (MV) multiplications. MV multiplications underpin much of modern computing and the imperative to carry out such calculations at ever-increasing speeds, but with lower energy consumption, is driving the development tensor processing units (TPUs). Why this is important: The team developed a new type of photonic TPU capable of carrying out multiple MV multiplications simultaneously and in parallel. The photonic processor is part of a new wave of light-based computing that could fundamentally reshape the digital world and prompt major advances in a range of areas, from AI and neural networks to medical diagnosis. [Click here to read on!]( [ML to Boost Particle Accelerator Diagnostics]( In brief: The U.S. Department of Energy's Thomas Jefferson National Accelerator has equipped operators of its primary facility, Continuous Electron Beam Accelerator Facility (CEBAF), with a new tool to help them rapidly address any issues that may arise. In preliminary tests, the tool successfully used ML to identify glitchy accelerator components in near-real-time. The CEBAF features a unique particle accelerator to explore the fundamental structure of nuclear matter. Powered by superconducting radiofrequency (SRF) cavities, CEBAF isn't immune from operational issues, therefore, a team of accelerator experts set out to build an ML system that could perform reviews in real-time. Their custom data acquisition system pulls information on cavity performance from a digital low-level RF system that is installed on the newest sections of a particle accelerator. The low-level RF system constantly measures the field in SRF cavities and tweaks the signal for each one to ensure optimal operation. Why this is important: The near-real-time feedback allowed CEBAF operators to make quick decisions on mitigating problems that arose in the machine during experimental runs. The idea is that eventually, the subject matter experts won't need to spend time looking at the data themselves to identify faults, allowing for greater amounts of time to be dedicated to physics research. [Click here to discover more!]( [Efficient Time-Series Analysis Using Pmdarima]( In brief: Time series analysis is one of the key concepts in data science. A time series is simply a series of data points ordered in time. In a time series, time is often the independent variable and the goal is usually to make a forecast for the future. There exist various forces that affect the values of the phenomenon in a time series. These are also the components of the time series analysis. These are: Secular trend or simple trend or long-term movement; seasonal variations, cyclical variations and; random or irregular variations. When conducting time-series analysis you may use either a univariate time-series analysis or multivariate time-series. This article explains the meaning of these terms and why pmdarima‘s auto_arima function is extremely useful when building an Auto Regressive Integrated Moving Average (ARIMA) model. Why this is important: This article in Towards Data Science demonstrates the efficiency of pmdarima’s auto_arima function compared to implementing a traditional ARIMA model. Information that will be of particular importance to those of us using Python and working as data scientists. [Click here to see the full picture!]( [AI Turns Descriptions into Actual Images]( In brief: A new AI by OpenAI allows you to create cute and bizarre images from scratch, based on your written description. Called DALL·E, the neural network is trained to generate images from text captions, using a dataset of text-image pairs. All you need to do is enter a description – the weirder the better – and wait for the results. While examples range from “an extreme close-up view of a capybara sitting in a field” to “an armchair in the shape of an avocado” the AI is particularly amusing when it comes to conjuring anthropomorphised versions of animals and objects in odd scenarios, such as the article’s example of an illustration of a baby daikon radish in a tutu walking a dog. “We find that DALL·E is able to create plausible images for a great variety of sentences that explore the compositional structure of language“ reads a description by OpenAI on its website. Why this is important: The skill behind the creation of these images is unquestionable. However, as data scientists, we must always consider ethical concerns. One of which is the potential for the AI to create complete false illustrations, which may contain real-life people or illegal images. [Click here to find out more!]( [SuperDataScience podcast]( In this week's [SuperDataScience Podcast](, we discuss the upcoming trends in 2021 such as training models without compromising privacy, measuring ROIs, racism and gender bias in AI, and much more! --------------------------------------------------------------- What is the Data Science Insider? This email is a briefing of the week's most disruptive, interesting, and useful resources curated by the SuperDataScience team for Data Scientists who want to take their careers to the next level. Want more conversations like this? Earlier this year, we held our first-ever DSGO Virtual Conferences, where more than 3,500 data scientists gathered to learn, grow, and connect! If you missed them or want to repeat this fantastic experience, stay tuned to our upcoming virtual and in-person events that will take your DS career to the next level. DSGO is your go-to place to elevate your technical skills, gain life-long career lessons from industry experts, and build lasting connections with data-driven peers. If you want to learn more and register for our future events, [click here](. Know someone who would benefit from getting The Data Science Insider? Send them [this link to sign up.]( # # If you wish to stop receiving our emails or change your subscription options, please [Manage Your Subscription]( SuperDataScience Pty Ltd (ABN 91 617 928) 131, 63 Blamey, St., Kelvin Grove, QLD 4059, Australia