Data Science Insider: December 1st, 2023

In This Week’s SuperDataScience Newsletter: Capitalism Triumphs in AI Vision Battle. Advancing Machine Unlearning. Mastering ARIMA: Key to Time Series Anomaly Detection. Mastering Data Science Soft Skills. Transforming Football Scouting, Coaching, and Game Strategy. 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. --------------------------------------------------------------- [DeepMind AI Discovers 2.2M Novel Materials]( brief: Google DeepMind researchers have employed the GNoME AI tool to uncover 2.2 million novel crystal structures, a breakthrough with implications for renewable energy and advanced computation which demonstrates the potential of AI in material discovery. Published in Nature, the findings present possibilities in renewable energy and advanced computation. The dataset, 45 times larger than historical discoveries, will offer 381,000 promising structures for testing. Ekin Dogus Cubuk, a co-author, emphasizes the impact on various technologies, stating, "It's hard to imagine any technology that wouldn't improve with better materials in them." This AI-driven approach accelerates materials science, presenting an order-of-magnitude expansion in stable materials and showcasing AI's ability to revolutionise experimental processes. Why this is important: The GNoME AI tool's ability to generate and assess novel crystal structures at an unprecedented scale not only showcases the power of AI in scientific exploration but also highlights its role in addressing grand challenges such as clean energy and environmental solutions. [Click here to learn more!]( [Revolutionising Self-Service Analytics: A Hierarchy Approach]( brief: In this comprehensive Towards Data Science article, Andrew Taft reflects on the evolution of self-service analytics from the 90s to the present, emphasizing the persistently low adoption rates despite technological advancements. Drawing inspiration from Maslow's Hierarchy of Needs, Taft introduces the "Self-Service Data Analytics Hierarchy of Needs." The hierarchy includes stages such as data collection, transformation, semantic layer creation, analysis, and, ultimately, self-actualization and transcendence. Taft underscores the importance of aligning people, processes, and tools in creating a data-driven organization, advocating for a collaborative, iterative approach. With a global self-service business intelligence adoption rate of 26%, this insightful framework aims to address the challenges hindering widespread adoption in the data analytics domain. Why this is important: Data scientists should recognize the significance of this framework as it provides a holistic approach to self-service analytics, acknowledging the interconnectedness of technological tools, organizational processes, and human factors. [Click here to read on!]( [Fourier Transform for ML]( In brief: The Fourier Transform, a mathematical technique pioneered by Jean-Baptiste Joseph Fourier in 1822, has transcended its roots in signal processing to become a cornerstone in ML, over recent years. This transformation dissects signals into constituent frequency components, facilitating analysis in the frequency domain. In ML, it proves indispensable across diverse applications. In time series analysis, it aids anomaly detection and trend analysis. In natural language processing, it enables frequency-based text analysis. Additionally, the Fourier Transform enhances feature engineering, optimizes Convolutional Neural Networks (CNNs) for image analysis, and supports data augmentation. This Medium article clearly outlines all of this and a provided Python code exemplifies its application in time series analysis. Why this is important: Whether unravelling temporal intricacies in finance or fine-tuning CNNs for image recognition, the Fourier Transform equips data scientists with a versatile toolset, enhancing model accuracy and performance. [Click here to discover more!]( [Top GitHub Alternatives for Data Scientists]( In brief: The KDNuggets article explores five alternatives to GitHub tailored for data scientists with unique needs. Kaggle stands out for its collaborative environment and real-world competitions, making it ideal for learning and skill enhancement. Hugging Face focuses on natural language processing, offering pre-trained models and a collaborative ecosystem for model training and sharing. DagsHub addresses the challenges of managing code, datasets, and ML models, emphasizing community collaboration. GitLab provides robust version control and project management tools, while Codeberg, a non-profit platform, emphasizes open source and privacy. Each platform caters to specific data science requirements, offering diverse options for collaboration, project management, and model handling. Understanding and utilizing these platforms can optimize collaboration, project management, and model development in data science endeavors. Why this is important: Data scientists, like you, would benefit from exploring these alternatives as they address specialized needs beyond GitHub and will enrich your performance. [Click here to see the full picture!]( [Crafting LinkedIn Profiles with AI Precision]( In brief: LinkedIn has introduced a generative AI feature, powered by OpenAI's GPT-4 model, to assist premium subscribers in crafting compelling headlines and "about" sections. While about 70% of users apply the AI-generated suggestions, feedback suggests that the content often feels robotic and may contain inaccuracies. Despite its potential utility for drafting initial profiles, users have cautioned against an overreliance on AI due to its cookie-cutter nature and inability to capture individual writing styles. LinkedIn has acknowledged the need for improvement, emphasizing ongoing efforts to enhance accuracy and tone. Data scientists should use the tool critically, ensuring profiles are well-filled for accurate results and considering manual edits for a more personalized touch. Why this is important: This development underscores the evolving integration of generative AI in professional platforms, offering both benefits and challenges for users seeking unique, authentic representation. [Click here to see the full picture!]( [Super Data Science podcast]( In this week's [SuperDataScience]( Podcast episode, Jon Krohn speaks to Mehdi Ghissassi, Director of Product Management at Google DeepMind, about the ethics and social impact of AI, keeping up with AI releases with safety in mind, and other pressing AI problems that keep him awake at night. In this episode, Mehdi and Jon also take a broader look at the current AI landscape, the opportunities for AI investors and startups, and what AI product managers need to get ahead. [Click here to find out 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 to take your data science skills to the next level? Check out the [SuperDataScience platform]( and sign up for membership today! 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), 15 Macleay Crescent, Pacific Paradise, QLD 4564, Australia