Harish Ojha’s Post

Join us for an interactive online workshop and a glimpse into the curriculum of our graduate programs in Decision Analysis. Led by Assistant Professor of Computational Sciences Aboozar Hadavand, in this session, we will discuss rigorous causal questions, the challenges of inferring causal relationships from real-world evidence, and the use of synthetic controls as a quantitative method of impact evaluation. You will walk away with the ability to understand cause-and-effect relationships, ask the right questions, and identify reliable and useful answers despite real-world constraints. Register here: https://lnkd.in/eTJ_uwxD #graduateschool #decisionmaking #workshop #causalinference #impactevaluation

You can't #solve any #problems you are unwilling to have... This was just one of the key takeaways from yesterday's Stanford Life Design Lab studio session, and while it may be tough to understand, its simplicity is profound. If you cannot accept something as a problem, you cannot solve it. Problem-solving is great, and our students learn that skill from P-16, but problem-finding is woefully undertaught in school. We provide the problems for students to solve, and sure, we give them alternative methods and teach them #CriticalThinking, but sourcing the problem is largely not our practice... In doing this, we eliminate their ability to be #creative and #curious, two essential components of high-level thinking. What problems are you not #accepting, halting your ability to solve? I can't wait to share what I learn in Day 2 later! Have a great Friday, all!

Join us for an interactive online workshop and a glimpse into the curriculum of our graduate programs in Decision Analysis. Led by Assistant Professor of Computational Sciences Aboozar Hadavand, in this session, we will discuss rigorous causal questions, the challenges of inferring causal relationships from real-world evidence, and the use of synthetic controls as a quantitative method of impact evaluation. You will walk away with the ability to understand cause-and-effect relationships, ask the right questions, and identify reliable and useful answers despite real-world constraints. Register here: https://lnkd.in/eTJ_uwxD #graduateschool #decisionmaking #workshop #causalinference #impactevaluation

🧠 𝐖𝐡𝐚𝐭 𝐢𝐬 𝐭𝐡𝐞 𝐑𝐨𝐥𝐞 𝐨𝐟 𝐒𝐦𝐚𝐥𝐥 𝐌𝐨𝐝𝐞𝐥𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐋𝐋𝐌 𝐄𝐫𝐚: 𝐀 𝐒𝐮𝐫𝐯𝐞𝐲 ✅ Scaling up #LLMs sizes results in exponentially higher computational costs and energy consumption, making these models impractical for academic researchers and businesses with limited resources. ✅ Small Models (#SMs) are frequently used in practical settings, although their significance is currently underestimated. This raises important questions about the role of small models in the era of LLMs. 🚀 This work systematically examined the relationship between LLMs and SMs from two key perspectives: Collaboration and Competition. 👉 Paper https://lnkd.in/dmDn9bkQ ♻️𝒌𝒊𝒏𝒅𝒍𝒚 𝒓𝒆𝒑𝒐𝒔𝒕♻️

We’re redefining adaptive learning with research-backed, personalized resources to address each student’s unique needs and give teachers the tools and time to build deeper relationships along the way. See us in action: http://spr.ly/6040io6yC

🎬 Unlock the Wonders of Science! with Keystone 🔬 📚 Why Science Experiments Spark Curiosity: 🔍 #Hands-On Learning: Engage directly with scientific concepts. ✨ #Visual Impact: See theories come to life before your eyes. 🧠 #Critical Thinking: Problem-solving in real-time. 🌟 #Endless Exploration: Every experiment leads to new questions. 🌍 Ready to ignite curiosity? Dive into a world of experiments and discover the excitement of learning. #ScienceExperiments #LearningThroughFun #CuriosityUnleashed #ScienceForStudents

Research Highlights (https://lnkd.in/gud9brZE): * Meta-heuristics are common, effective approaches across various optimization problems * Although hybrid meta-heuristics are the best-performing methods in many problem fields within OR, their detailed study is commonly overlooked in the literature * This work provides a comprehensive analysis of meta-heuristics and a classification for hybrid meta-heuristics in educational timetabling * This is a great read for those working on scheduling or (hybrid) meta-heuristics Feel free to share your thoughts or ask questions in the comments. If you found this research insightful, please help it reach a wider audience by liking and reposting this post. Thanks! #elsevier #operationsresearch #artificialintelligence #optimization #combinatorialoptimization #scheduling #metaheuristic #educationaltimetabling #universitycoursetimetabling

🔬📚 Build a strong foundation in Science with these essential study tips! 🧠 Master the basics, stay curious, and practice regularly to boost your skills. 🚀 #ScienceStudyTips #StrongFoundation #BenchMarkLearning #StudySmart

Embracing Rapid Experimentation for Authentic Scientific Inquiry “Planning and carrying out investigations doesn’t have to be an exhaustive process that consumes time and resources. In fact, rapid experimentation can be a powerful approach, allowing students to explore and understand scientific phenomena quickly and effectively. The essence of experimentation lies in offering explanations for observed phenomena, rather than meticulously detailing every variable. By embracing rapid experiments, students can engage directly with the scientific process, form hypotheses, test them in real-time, and draw meaningful conclusions—all within a manageable timeframe. For teachers facing time constraints or limited resources, shifting to a rapid experimentation model can be a game-changer. This method emphasizes the core of scientific inquiry: curiosity, observation, and critical thinking. In contrast, relying solely on simulations and teacher-centered demonstrations can hinder students from grasping the real meaning of inquiry. Textbook experiments often fail to provide the hands-on experience that is crucial for deep understanding. Whether it’s a quick in-class experiment using everyday materials or a simplified version of a more complex investigation, the goal remains the same—to spark understanding and foster a hands-on approach to learning. Remember, the key is not in the complexity of the experiment but in the clarity of the insights it provides. #ScienceEducation #RapidExperimentation #HandsOnLearning #InquiryBasedLearning #STEM #TeachScience #ActiveLearning

Grateful I achieved a perfect score on a challenging project in the "Classification and Hypothesis Testing" course, part of the MIT IDSS (MIT Institute for Data, Systems, and Society) program. 📊 Project Overview: The project involved analyzing data for ExtraaLearn, an EdTech startup, to identify potential paying customers from their leads. This required building and fine-tuning machine learning models to understand and predict customer behavior. 🚀 Key Takeaways & Scores: Exploratory Data Analysis (6/6): Successfully defined the problem, performed comprehensive univariate and bivariate analyses, identified outliers, and used visualizations to uncover patterns and insights. Data Pre-processing (4/4): Excelled in data cleaning, feature engineering, and preparing the dataset for modeling, ensuring high data quality and relevance. Model Building - Decision Tree (5/5): Built an accurate Decision Tree model, providing clear explanations of its performance. Model Building - Random Forest (5/5): Developed a robust Random Forest model, with insightful commentary on its performance. Model Performance Evaluation (10/10): Expertly tuned model hyperparameters using GridSearchCV, evaluated models with appropriate metrics, and analyzed feature importance. Actionable Insights & Recommendations (5/5): Offered valuable insights and practical recommendations for ExtraaLearn, based on thorough analysis. Presentation Quality (5/5): Maintained a high standard in report writing, with clear, well-commented code and insightful observations. 👨🏫 Professor's Remarks: "I'm truly impressed by the quality of your work. Your explanation is clear and thorough. You've effectively utilized both decision tree and random forest models, delved into feature importance, and provided actionable insights. Your dedication is commendable. To further refine your skills, I recommend concentrating on the dataset's numerous features, delving into advanced preprocessing techniques, and developing more robust algorithms like Xgboost and Catboost. Keep up the excellent work!" This journey has been incredibly enriching, blending theoretical knowledge with practical application. I'm grateful for the opportunity to learn and apply these skills in real-world scenarios. Looking forward to continuing this learning journey! #DataScience #MachineLearning #MITIDSS #EdTech #Analytics #Achievement #ContinuousLearning