Javed Alam’s Post

If you're like me and sometimes run into GPU limitations on your local machine, I Tried a neat workaround to run Ollama in Google Colab, allowing you to leverage Colab's resources for your AI projects. You can find how To set it up Here : https://lnkd.in/deU6-HHY #GenerativeAI #LLM #RAGsystems #ArtificialIntelligence #MachineLearning #DataScience #AIResearch #TechInsights #AIApplications #AICommunity #Ollama #Python

🚀 Built an ANN for Regression I recently developed an Artificial Neural Network to tackle a regression problem. With optimized layers and hyperparameters. 🔍 Tech Stack: Python, TensorFlow, Colab 📊 See the Notebook: https://lnkd.in/g9KACMjP Have you tried ANNs for regression? Let’s connect and share insights! #MachineLearning #NeuralNetworks #Regression #AI

🚀 Exploring Non-Euclidean Clustering Algorithms! 🌐 Excited to share my latest project on clustering in non-Euclidean spaces using spherical and hyperbolic distance metrics. Dive into the code and visualizations to see how we can push the boundaries of traditional clustering methods. Check it out on Colab: https://lnkd.in/gY-D6YF3 #DataScience #MachineLearning #Clustering #NonEuclidean #AI #Python

Notebook: https://lnkd.in/gyXaJQVv 💀 Proud to present my most recent project: A Python-Based Global Search Optimization Platform (GSOP) That Leads the Industry! 📊 This state-of-the-art notebook features: ✅ Modular design incorporating multiple optimization algorithms ✅ Parallel processing for improved performance ✅ Extensive statistical analysis and benchmarking ✅ Interactive visualizations and experiment configuration ✅ Model persistence for simple sharing and replication This tool can greatly enhance your optimization workflows, regardless of your field of expertise—data science, machine learning, or operations research. Look over the notebook and let me know what you think! How may you use this in your professional life? #MachineLearning #DataScience #Optimization #Python #OpenSource 🐉 If you think this is helpful, please like, comment, and share!

It was claimed on this forum that Large Language Models will not solve optimization problems "any time soon" (see thread: https://lnkd.in/eBJXufdF). I guess they won't if nobody asks them to. But, since I'm paying the $200 anyway, and since Steve Moffitt if no-one else was interested, I might as well get this little research project started with this prompt: "Take all the methods in HumpDay at https://lnkd.in/gr6shcD and understand the ideas in them. Then devise a new heuristic derivative-free optimization procedure that combines some ideas. Then benchmark it in Python." Here is the result: https://lnkd.in/eDnttCr5 self-described as "A toy meta-heuristic that adaptively selects among sub-optimizers". The code runs first time. I'm not suggesting this is publishable research, or even logical, just that somebody might have fun running with this one given that that there are reasonably straightforward ways of evaluating the thousands of candidate algorithms that can be generated in this manner. If you do co-invent something sneaky, I'd love to accept your pull request: https://lnkd.in/eAm-V2qt #llms #optimization https://lnkd.in/eDnttCr5

Lectures 10, 11, and 12 Highlights: Advanced AI with Sir Irfan Malik As-salamu alaykum! Today, in our Advanced AI Learning Journey, we are covering three essential lectures with Sir Irfan Malik. In these sessions, we will discuss: - Conditional statements - Loops (while loops and for loops) - User-defined functions in Python - Open Ai To enhance your learning experience, we've prepared a Colab notebook with practical exercises. Check out the link below and follow along with the exercises. 📚 Colab Notebook Link of Lecture 10 https://lnkd.in/d_fNBGFy 📚 Colab Notebook Link of Lecture 11 https://lnkd.in/dA67vN-P 📚 Colab Notebook Link of Lecture 12 https://lnkd.in/dEfsjd5v Let's continue building our Python skills together! #AI #Python #Functions #Loops #ConditionalStatements #LearningJourney #ZeonSolutions #TechSkills #ContinuousLearning

# LLM-MICRGRD-2 🚀 Understanding Derivatives: A Practical Example in Python In mathematics, derivatives help us understand how a function changes with respect to its input variables. Let’s explore derivatives with respect to the variables of a simple function: d=a⋅b+c In this example, we’ll compute the derivatives of d with respect to a, b, and c, using Python and a numerical approximation method (finite difference). 🎯 Goal: We want to calculate the derivative of d for each variable: With respect to a With respect to b With respect to c 🛠 Python Code to Calculate Derivatives: Try it out yourself https://lnkd.in/dh7zVyz4 🔍 Key Results: Derivative of d with respect to a: ∂a/∂d=∂a/∂(a⋅b+c)=b By changing a slightly, we find that the derivative is b, which in this case is -3.0. This means that for every unit change in a, d changes by -3.0 units. Derivative of d with respect to b: ∂b/∂d=∂b/∂(a⋅b+c)=a By changing b slightly, we find that the derivative is a, which in this case is 2.0. This means that for every unit change in b, d changes by 2.0 units. Derivative of d with respect to c: ∂c/∂d=∂/c∂(a⋅b+c)=1 The derivative with respect to c is always 1, meaning that for every unit change in c, d changes by exactly 1 unit. 🔍 Summary of the Derivatives: ∂d/∂a= b ∂d/∂b= a ∂d/∂c= 1 🌟 Takeaway: This example shows how derivatives are used to understand the rate of change of a function with respect to its variables. In machine learning and optimization, we often need to compute derivatives (or gradients) to adjust model parameters effectively. 💡 Pro Tip: Derivatives are essential for understanding and optimizing mathematical models, especially in machine learning algorithms like gradient descent. #DataScience #MachineLearning #Python #Derivatives #Mathematics #Optimization #TechTips #PythonProgramming

🚀 Excited to share my latest project on Multi-dimensional Knot Theory! 🧵 🔍 Project Highlights: * Persistent Homology: Leveraging Gudhi library for topological data analysis. * Machine Learning: Classifying knots using a neural network model. * Visualization: Detailed persistence diagrams for in-depth analysis. Check out the full project and code on Google colab! #DataScience #MachineLearning #Topology #KnotTheory #Python

Lately, I've been reading a lot about creating Neural Networks with Python and NumPy. The whole process has been fascinating and informative. Below is a Neural Network solving the classic XOR gate problem:

📚 New notebook alert! Build Agentic RAG using Hugging Face's smolagents library and compare it to Vanilla RAG. Tech stack: • Unstructured Platform for PDF processing • DataStax AstraDB for vector storage • new `smolagents` library for agent implementation • OpenAI models for embeddings & generation LLM Learn how to: • Process PDFs with Unstructured Platform & store in DataStax AstraDB • Build Vanilla RAG from scratch in Python • Create Agentic RAG using smolagents and different types of Agents • Improve answer quality through multi-step retrieval 🔗 https://lnkd.in/ee9UHScb