VECTOR Inc.’s Post

🚀 Project Completed: ASL Digit Recognition with CNN 🚀 I'm thrilled to share that I've successfully completed my project on American Sign Language (ASL) digit recognition using Convolutional Neural Networks (CNN) as part of my Artificial Intelligence Project Based Learning Subject! 📚 Project Overview: The goal of this project was to develop a machine learning model capable of recognising ASL digits. By leveraging the Sign Language MNIST dataset, I implemented a CNN that accurately identifies ASL digits, making significant strides in the field of gesture recognition and accessible technology. ✨ Key Features: Data Preprocessing and Augmentation Model Building and Training with CNN Hyperparameter Tuning Class Weight Adjustment Performance Evaluation (Accuracy, Precision, Recall, F1 Score) 🔧 Technologies Used: Python TensorFlow/Keras Scikit-learn Pandas & NumPy 🙏 Special Thanks: I would like to express my sincere gratitude to: Syed Karar Haider Bukhari (Project Supervisor) for your invaluable guidance and support throughout the project. Osama Imran (Project Coach) for your insightful advice and encouragement. Your expertise and teamwork were instrumental in the successful completion of this project. I couldn't have done it without you! #ArtificialIntelligence #MachineLearning #DeepLearning #ASL #GestureRecognition #ConvolutionalNeuralNetwork #AIProjects #TeamWork #Gratitude #LearningJourney

Excited to announce the successful completion of my Artificial Intelligence training at Zewail City, accredited by the International Accreditation Organization (IAO) , ISO - International Organization for Standardization and Zewail City of Science and Technology ! 🎓 Here's a glimpse into what I've learned: 🔹 Delved into the fundamentals of AI and its diverse applications. 🔹 Mastered Python programming for AI development. 🔹 Explored essential data structures, algorithms, and programming concepts. 🔹 Dived deep into Machine Learning algorithms and techniques. 🔹 Ventured into the realm of Deep Learning, including Neural Networks and NLP. 🔹 Explored Reinforcement Learning and its applications. 🔹 Got hands-on experience with Embedded Systems in AI and Big Data Analytics. 🔹 Explored modern roles for Logic in AI and identified vulnerabilities. 🔹 Discussed ethical concerns and the future of AI. Excited to leverage this knowledge to make meaningful contributions in the field of AI! #ArtificialIntelligence #MachineLearning #DeepLearning #AIethics #ZewailCity #IAOcertified ✅✨

Attended a 2-Day Workshop on "Machine Learning and Neural Networks" by MathLab SkillRealm , conducted by Petrichor '25 of IIT Palakkad on September 28th & 29th, 2024. This workshop provided a valuable, hands-on experience in Machine Learning and an introduction to Neural Networks. The sessions were led by the wonderful mentors, Mohammed Hashim and Hanna Maryam and other core team mentors. Day 1: Foundations of Machine Learning - Started with Python practice questions to reinforce coding skills. - Covered core topics: - The history of Artificial Intelligence. - Types of Machine Learning and their applications. - Key concepts like Linear and Logistic Regression, Learning Rate, and Iteration. Day 2: Coding and Neural Networks - Hands-on coding to develop models. - Overview of Neural Networks and creating an AI model, which included: - Idea generation. - Mathematical modeling. - Python code implementations and Clustering techniques. This workshop was an excellent opportunity to connect theoretical insights with practical applications, enhancing my understanding of ML. #artificialintelligence #machinelearning #neuralnetworks

I’m very happy to share that l've crossed the finish line of the Large Language Models Bootcamp by SDAIA | سدايا and have successfully passed the final examination of the Large Language Models Practitioner Certificate.🌟 During this intensive program, l gained valuable knowledge in ML Deep Learning and Backpropagation, Neural Network with TensorFlow and Keras, Deep Learning for Image Data, Text Data Analysis, Understanding LLMS, Advanced Concepts and Optimization Techniques for LMS, Advanced Language Generation, State-of-the-art LLMs Collaborating. with talented peers and working on real-world projects enhanced my skills and problem-solving abilities. Looking forward to applying my new LLM skills to real-world challenges. 🚀🔍💡 #SDAIA #AI #DataScience #LanguageModels #Artificiallntelligence

Thrilled to announce our paper, "Evaluate effect of 126 pre-processing methods on various artificial intelligence models accuracy versus normal mode to predict groundwater level (case study: Hamedan-Bahar Plain, Iran)," is now published in Heliyon Journal. The main objective of this study is to compare the effect of various preprocessing methods on input data for different artificial intelligence (AI) models to predict groundwater levels (GWLs). Firstly, 126 method of data pre-processing was done by python programming which are classified into three classes: 1- statistical methods, 2- wavelet transform methods and 3- decomposition methods; later, various pre-processed data used by four types of widely used AI models with different kernels, which includes: Support Vector Machine (SVR), Artificial Neural Network (ANN), Long-Short Term memory (LSTM), and Pelican Optimization Algorithm (POA) - Artificial Neural Network (POA-ANN) are classified into three classes: 1- machine learning (SVR and ANN), 2- deep learning (LSTM) and 3- hybrid-ML (POA-ANN) models, to predict groundwater levels (GWLs). You can find the paper at the following link: https://lnkd.in/d7zYz7tC #Artificial_Intelligence #Machine_Learning #Deep_Learning #Hybrid_Algorithm #Decomposition_Methods #Groundwater

🚀 Excited to share my recent deep dive into binary classification using a synthetic dataset with the power of PyTorch and visualized with Matplotlib! 🚀 I tackled a classic binary classification problem with the make_circles dataset from sklearn. This dataset is a perfect playground to demonstrate the effectiveness of linear and non-linear layers in neural networks. Key Steps: 1.Data Preparation: Generated a concentric circles dataset using make_circles, which provides a challenging non-linear problem. 2.Model Architecture: Built a neural network using PyTorch with a combination of linear and non-linear layers. The architecture included: * Linear layers to transform inputs * Non-linear activation functions, specifically Sigmoid and ReLU, to introduce non-linearity and capture complex patterns. 2.Training: Trained the model using PyTorch's robust training loops, leveraging stochastic gradient descent for optimization. 3.Visualization: Utilized Matplotlib to visualize decision boundaries and data distributions, highlighting the model's ability to separate the classes effectively. Tools Used: PyTorch: For building and training the neural network Matplotlib: For visualizing the data and decision boundaries sklearn: For generating the make_circles dataset Curious to see the code and visuals? Drop a comment or message me! 🌟 #MachineLearning #PyTorch #DataScience #DeepLearning #AI #Visualization #NeuralNetworks #BinaryClassification #Python

Leveraging Deep Learning for macroeconomic forecasting 📊 Intrigued by the potential of deep learning in macroeconomic forecasting, I explored the use of Long Short-Term Memory (LSTM) models. LSTMs are powerful neural networks adept at capturing long-term dependencies within time series data, making them well-suited for this field. To this end, I created a user-friendly Colab notebook in Python designed to fit an LSTM model on the latest FRED-MD dataset, which is a comprehensive collection of over 100 US macroeconomic time series starting from 1959 (https://lnkd.in/dewkr98j). The notebook offers flexibility by allowing users to: 1️⃣ Specify the target variable to forecast; 2️⃣ Set the desired forecast horizon; 3️⃣ Customize the model architecture, even beyond LSTM. Additionally, it boasts the potential for importing user-defined datasets, enabling broader applications. I used the notebook to fit a model aimed at generating 3-month ahead forecasts for US real personal income (PDF attached). Here is the link to the notebook👇 https://lnkd.in/dUujgXVu Any feedback is appreciated! #deeplearning #machinelearning #LSTM #forecasting #macroeconomics

I’m very happy to share that l've crossed the finish line of the Large Language Models Bootcamp by SDAIA | سدايا and have successfully passed the final examination of the Large Language Models Practitioner Certificate.🌟 During this intensive program, l gained valuable knowledge in ML Deep Learning and Backpropagation, Neural Network with TensorFlow and Keras, Deep Learning for Image Data, Text Data Analysis, Understanding LLMS, Advanced Concepts and Optimization Techniques for LMS, Advanced Language Generation, State-of-the-art LLMs Collaborating. with talented peers and working on real-world projects enhanced my skills and problem-solving abilities. Looking forward to applying my new LLM skills to real-world challenges. 🚀🔍💡 #SDAIA #AI #DataScience #LanguageModels #Artificiallntelligence

Just came back from a workshop that I'm organizing together with the VDI Wissensforum GmbH. I had a great time, learned a lot and it made me realize (once again) how much has changed in AI over the last 5 years: When we started with the first workshop in 2019, the format 💠 targeted engineers who were able to code in Python 💠 focussed on the convolutional neural networks architecture 💠 involved building neural networks from scratch in Keras 💠 emphasized use cases that cost >100k to implement On Tuesday and Wednesday, we ✅ were a diverse group with little to no Python knowledge ✅ discussed Transformers (briefly) and RAG (in detail) ✅ used off-the shelve models and tools ✅ talked about use cases from 2k to 200k To me this is another testament to how much more accessible and ubiquitous AI has become. If you are interested in learning how to use retrieval augmented generation in industrial applications check out the workshop: https://lnkd.in/eTsG3x6h Thanks Nicolas Regiani for creating this together! #vdi #rag #industrialai

🚀 Excited to share my latest project: IMDB Movie Review Sentiment Analysis with Simple RNN! 🎬 I built an end-to-end deep learning application that classifies movie reviews as positive or negative using a Simple Recurrent Neural Network (RNN). The project involves training the model on the IMDB dataset and deploying it as an interactive web application with Streamlit. 🔗 GitHub Repository: https://lnkd.in/dD6H45Cr 🌐 Web Application: https://lnkd.in/dUeeupU7 Key Features: * Preprocessing and padding of IMDB movie reviews * Training a Simple RNN model with early stopping to prevent overfitting * Saving the trained model and deploying it with Streamlit for real-time sentiment analysis * Interactive interface for users to input movie reviews and receive sentiment predictions with confidence levels This project has been an incredible learning experience, and I’d like to extend my gratitude to Krish Naik for the valuable tutorials and insights that helped me along the way. Feel free to check out the project on GitHub and try out the web application. Your feedback and suggestions are always welcome! #DeepLearning #MachineLearning #ArtificialIntelligence #NLP #SentimentAnalysis #Python #TensorFlow #Streamlit #IMDB #MovieReviews #DataScience