Bioengineering’s Post

Updated !! My ongoing study as part of the Digital Talent Incubator program in Data Science and Machine Learning at Purwadhika Digital Technology School! I'd like to discuss the AMAZING advances in machine learning in the chemical sector. Machine learning is revolutionizing the way we approach chemical research and development, particularly in the design and discovery of heterogeneous catalysts. These catalysts are critical for accelerating chemical reactions in a variety of industrial processes, and using ML can greatly improve our understanding and efficiency in this domain. My most recent topic investigates how machine learning accelerates active site determination and identifies important descriptors and trends, etc. By incorporating ML approaches, we can achieve major advances in catalyst design, resulting in more efficient and effective industrial operations. Thank you, Mr. Mohammad Digjaya, for guiding me through the chemistry of machine learning. You can also obtain your chemistry right here (https://lnkd.in/gCwk2E4h). I hope we have some chemistry—in machine learning. #MachineLearning #DataScience #Chemistry #Catalyst

🚨 Call for Papers: Special Issue on Machine Learning in Mechanical and Materials Science 🚨 I’m excited to announce that I’m serving as a Guest Editor for a special issue in the SCI-indexed journal Computer, Materials & Continua (Impact Factor: 2.0). This is an open-access journal, and we are offering free-of-cost publication in this special issue! Submissions are now open! This issue will focus on cutting-edge research at the intersection of Materials Science and Machine Learning, with topics including: 🔹 Materials Discovery 🔹 Predictive Modeling 🔹 Data Mining 🔹 High Entropy Alloys 🔹 Material Optimization 🔹 Neural Networks in Materials 🔹 AI-driven Materials Design 🔹 Genetic Algorithms in Materials 🔹 Sustainable Materials 🔔 Submission Deadline: July 31, 2025. We welcome contributions from researchers and practitioners in these areas. Don’t miss this chance to share your work with the global research community and publish free of charge! For submission details and more information, please visit: https://lnkd.in/gjyP3fPK #MaterialsScience #MachineLearning #AI #MaterialsDiscovery #PredictiveModeling #HighEntropyAlloys #NeuralNetworks #SustainableMaterials #Research #OpenAccess #FreePublication #SpecialIssue #Innovation

Machine Learning in materials science is a lot like having a million grad students plugging away at material datasets. Sure, they can handle a lot of ones and zeros, but they don’t really know the full story, and sometimes the results can meander from the intended aims and goals. When the group of students is small, it is easy to manage. When the group gets larger, it becomes more of a challenge. With enough effort and data to process, results can be extracted. I’ve discussed this before in an editorial – so-called DICO for “data in, correlation out” (see Link 1 below). Given a large enough data set, one can always find a chain of correlations (negative or positive) between subsets of data. I.e., the computer will spit out a theoretical solution (or material), but will it translate to a physical solution (or material)? E.g., How can we ensure the results are meaningful? The February Issue of Matter has two different papers that help tackle this problem. The first work by Phillip M. Maffettone and colleagues (Link 2) from Brookhaven National Laboratory uses Bayesian optimization to emulate the perspective of experts driving an experiment. In effect, it asks the question, “If a positive/desired result is computationally found, can that target be experimentally achieved?” Rather than crunch the materials data alone for potential solutions (depending on desired objectives), experimental constraints are introduced via clever scientific value functions (SVFs) to guide the optimization. In effect, these SVFs can mirror the perspective or actions of human experts. The second work by Gilad Kusne colleagues (Link 3) from National Institute of Standards and Technology (NIST) introduces human input into an autonomous experimentation (AE) workflow, to subjectively push the optimization by expertise, intuition, and/or curiosity. Human input can directly affect the next material domain investigated, enabling the power of computational efficiency with the flexibility of human creativity. This approach transforms ML/AI from a blackbox to more of a user-guided instrument. The tradeoff is a loss of some autonomy (and efficiency), but the gain of ML trustworthiness and interpretability.   Both works see a need to manage pure data processing with more intuitive methods, providing a human touch in the digital space, helping to ground virtual outcomes with reality. #MachineLearning #ArtificialIntelligence #HumanInTheLoop #MaterialsScience   Link 1 – Cranford, Editorial, “Vitriol, brimstone, and DICO in materials science” https://lnkd.in/edRDBkMw Link 2 – Maffettone and colleagues, “Flexible formulation of value for experiment interpretation and design” https://lnkd.in/eGzVc_V6 Link 3 – Kusne and colleagues, “Human-in-the-loop for Bayesian autonomous materials phase mapping” https://lnkd.in/etFAXMfR

Here is my blog article on how machine learning can be used for material science research. Use cases with codes and explanation are provided for effective learning

🤔 𝐓𝐢𝐫𝐞𝐝 𝐨𝐟 𝐜𝐨𝐩𝐲𝐢𝐧𝐠 𝐫𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐩𝐚𝐩𝐞𝐫𝐬? 𝐋𝐞𝐚𝐫𝐧 𝐭𝐨 𝐀𝐧𝐚𝐥𝐲𝐳𝐞 𝐭𝐡𝐞𝐦 𝐥𝐢𝐤𝐞 𝐚 𝐁𝐎𝐒𝐒! Research papers are the foundation of advancements in Machine Learning, but their technical format and structure can be intimidating. This makes it difficult for engineering students to grasp key concepts and apply them to real-world problems. Register here : 👇🏼 https://lnkd.in/gaUfVuQq 📍 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐇𝐔𝐁 𝐢𝐬 𝐩𝐫𝐞𝐬𝐞𝐧𝐭𝐢𝐧𝐠 𝟒-𝐰𝐞𝐞𝐤 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 𝐖𝐨𝐫𝐤𝐬𝐡𝐨𝐩 𝐞𝐦𝐩𝐨𝐰𝐞𝐫𝐬 𝐘𝐎𝐔! In this 4-week workshop, you'll learn how to analyze research papers like an expert. You'll gain the skills you need to: • Understand the structure and purpose of research papers • Develop effective analysis strategies • Analyze specific research papers • Identify the key strengths and limitations of research papers ❗ The workshop is designed for engineering students with little to no experience with machine learning. Starting Date : 28 April #research #papers #ml #AI #workshop

🎓 I am proud to announce my graduation from the MIT Schwarzman College of Computing's Applied Data Science and Machine Learning program. This transformative journey has significantly deepened my understanding of data-driven decision-making. The program offered me the chance to explore advanced techniques such as predictive analytics, natural language processing, neural networks, and deep learning. These tools empower me to not only fundamentally understand a key component of our business at M42 Health but also consider pursuing a career in this field if I choose. For my capstone project, I developed a recommendation system utilizing 780 TB of product and customer preference data, further equipping me to partially understand the complexity of what our Applied Data Science and Artificial Intelligence team at M42 Health go through and be able to adequately support them as we tackle complex problems in healthcare. I'd like to extend a heartfelt thank you to the staff at MIT and Great Learning, as well as my family and friends. A special shoutout to my bestie—your snarky comments have been part of my success. Balancing a full-time professional role at an ambitious company, being a husband and father of three, an active member of the fitness circuit, and studying was not easy. Your unwavering support made this possible, and I am eternally grateful. Next on my agenda: GenAI with Microsoft! "We only stop learning when we stop breathing." - Unknown #DataScience #MachineLearning #MIT #LifelongLearning #HealthTech

In the middle of my lab transition at MIT, there were days when I had nothing to do. I had quit my original lab, but not found a new one yet. My days had no specific agenda. I was just exploring. Exploring at MIT is one of the best things you can do. (1) For a week, I went to the MIT Library and read Ian Goodfellow’s Deep Learning book. (2) One week, I walked to the Harvard Bookstore from MIT every single day. I browsed through books on artificial intelligence and made notes. (3) I used to attend every seminar at MIT, and ask questions to the speaker. This exploration led me to discover Scientific Machine Learning. I attended a seminar by Chris Rackauckas which showed how we can combine physics with machine learning. That seminar was like lightning had struck me and suddenly I had found a new path. 4 reasons why that seminar such an eye opener for me: (a) I was finding it very hard to transition to machine learning (ML). I had always been a mechanical engineer and loved differential equations and physics. (b) All ML courses I took always did toy projects like housing price predictor, movie review analyser, tweet sentiment predictor etc. I could not relate with these projects. I wanted to learn ML, but still remain connected to engineering, physics and differential equations . (c) That talk on Scientific Machine Learning showed me that ML and engineering need not be two separate fields. We can merge both of them and create powerful projects. (d) That way, I could merge my domain engineering knowledge with machine learning! After that workshop, I was curious to apply machine learning to the fluid mechanics research problem in my original lab. That curiosity led to a passion, and I ended up switching fields to Scientific ML. I joined a new research group at MIT with Chris Rackauckas as my advisor. For the next 3 years, Scientific ML changed my life. It led to multiple publications, a PhD thesis at MIT and multiple job offers. I still continue teaching Scientific ML to engineers all over the world. The field which changed my life will always remain my passion. If I had not been in that exploration phase at MIT, I would have never attended that Scientific ML seminar. Maybe I would have quit my PhD at MIT. Sometimes it's better to not have everything figured out. You can use that time to explore. That exploration may lead you to discovering your true passion! P.S: We are conducting a free webinar on how to transition to ML for all engineers. Registration link in the comments. 

🔬 Unveiling the Alchemy of Machine Learning in Chemistry 🌐💡 In the intricate world of chemistry, machine learning emerges as a transformative force, revolutionizing the way researchers explore and innovate. Let's dive into some key applications: Predicting Compound Properties: Machine learning steps in to predict crucial properties of chemical compounds—binding energy, acidity, and activity in biological systems. This empowers researchers to swiftly assess new compounds, optimizing them for specific applications with efficiency and precision. Design of New Compounds: Leveraging machine learning algorithms, researchers can identify patterns in the structure and properties of compounds. This knowledge becomes the cornerstone for designing novel compounds with desired properties, ushering in a new era of targeted compound development. Pattern Recognition: Machine learning's prowess in pattern recognition becomes indispensable for identifying anomalies and unusual patterns in chemical data, such as spectra and chromatograms. This capability aids in pinpointing specific compounds or reactions. Text Analysis: Machine learning extends its reach into textual data, analyzing articles and publications to identify keywords and terms related to chemical processes or compounds. This keeps researchers abreast of the latest trends, fostering a dynamic environment of continuous learning. 🧪 Empowering Chemistry Research: Machine learning stands as a robust tool for chemistry researchers, providing a lens to analyze complex and nonlinear data. Through the deployment of machine learning algorithms, researchers unearth hidden patterns that lead to new discoveries and deeper insights into chemical processes. The synergy of machine learning and chemistry not only expedites research but also opens doors to innovation, where data-driven insights catalyze groundbreaking advancements. Let's continue to explore the boundless possibilities at the crossroads of technology and chemistry! 🌐⚗️ #MachineLearning #Chemistry #TechInnovation #DataAnalysis #ResearchAdvancements

Thrilled to share I've completed the 18-hour "Introduction to Machine Learning" on Coursera, achieving a perfect 100% score. As an electrical engineering professional, I'm excited to seamlessly integrate these skills with AI and ML, bridging gaps and expanding my expertise. Ready to apply these learnings to real-world projects. #MachineLearning #AI #ElectricalEngineering #ContinuousLearning

📘 Exciting Announcement: I am thrilled to share that We have recently published a new book under Nirali Publications tailored specifically for BE AI&DS students. Titled "Distributed Computing," this book aims to provide a comprehensive and accessible resource for understanding the intricacies of distributed systems. It covers essential topics crucial for students pursuing advanced studies in Artificial Intelligence and Data Science, offering practical insights and theoretical foundations alike. I'm delighted to contribute to the academic community and provide a valuable learning tool for future technologists. Explore more about the book and how it can benefit students in their educational journey! #DistributedComputing #AI #DataScience #Education #NewBook