987 results
987 results
Explore courses, coaching and communities alongside independent reviews and available community feedback.
Pytorch for deep learning (udemy)
0.0
PyTorch for Deep Learning is a comprehensive course that teaches deep learning using PyTorch, one of the most popular deep learning frameworks. This hands-on course covers PyTorch fundamentals, tensor operations, automatic differentiation, neural network construction, and advanced deep learning techniques. Students learn to build and train various neural network architectures including feedforward networks, convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers. The curriculum includes PyTorch basics and tensor manipulation, building custom datasets and data loaders, implementing gradient descent from scratch, creating neural network architectures with torch.nn, transfer learning techniques, computer vision applications, natural language processing with transformers, and deployment strategies. Key projects include image classification systems, object detection models, sentiment analysis applications, and generative models. The course emphasizes practical implementation with real-world datasets and includes best practices for model training, debugging, and optimization. Students learn to use PyTorch Lightning for production-ready code, implement custom loss functions, and handle common deep learning challenges. Advanced topics include attention mechanisms, adversarial training, and model interpretability. Duration: 40+ hours of video content, self-paced with lifetime access. Prerequisites: Python programming experience and basic machine learning knowledge recommended. Includes downloadable resources, coding exercises, and certificate of completion.
Elements of ai (university of helsinki)
0.0
Elements of AI is a free online course designed to introduce artificial intelligence to everyone, regardless of background or technical skills. Created by the University of Helsinki in collaboration with Reaktor, this course demystifies AI and makes it accessible to a broad audience. The curriculum covers what AI is and is not, real-world applications of AI, machine learning basics, neural networks fundamentals, implications of AI for society, and ethical considerations in AI development. The course emphasizes understanding rather than programming, making it perfect for non-technical learners who want to understand AI concepts. Key topics include problem-solving with AI, uncertainty and probability in AI systems, optimization algorithms, the Bayes rule, classification and clustering, and the philosophy of AI. Students learn through interactive exercises, real-world examples, and case studies rather than complex mathematics. The course addresses common misconceptions about AI, discusses current limitations and capabilities, and explores future possibilities. No programming experience required, making it accessible to business professionals, students, and curious individuals. The course has been completed by over 1 million people worldwide and is available in multiple languages. Duration: 6 weeks, 2-4 hours per week, completely self-paced. Prerequisites: none - designed for absolute beginners. Completely free with optional certificate.
Computer vision nanodegree (udacity)
0.0
The Udacity Computer Vision Nanodegree is an intensive program designed to teach the fundamentals and advanced techniques of computer vision and image analysis. This hands-on program covers image processing fundamentals, feature detection and matching, object tracking, facial keypoint detection, and deep learning for computer vision. Students work with OpenCV, PyTorch, and other industry-standard tools to build real-world computer vision applications. The curriculum includes image representation and analysis, edge detection, Harris corner detection, histogram of oriented gradients (HOG), scale-invariant feature transform (SIFT), convolutional neural networks for image classification, transfer learning, object detection with YOLO and R-CNN, semantic segmentation, and generative adversarial networks (GANs) for image synthesis. Key projects include building a facial keypoint detection system, implementing automatic license plate recognition, creating an image captioning system, and developing a landmark detection and tracking system. The program emphasizes practical applications including autonomous vehicles, medical imaging, surveillance systems, and augmented reality. Students learn to preprocess images, extract meaningful features, train deep neural networks, and deploy computer vision models in production environments. Duration: 4 months at 10-15 hours per week. Prerequisites: intermediate Python programming and basic machine learning knowledge. Includes 1-on-1 mentoring, career services, and job placement assistance.
Aws machine learning specialty certification
0.0
The AWS Machine Learning Specialty Certification validates expertise in building, training, tuning, and deploying machine learning models on AWS. This certification covers the full ML lifecycle including data preparation, feature engineering, model training, evaluation, and deployment using AWS services. The curriculum includes AWS ML services overview, data engineering for ML with services like S3, Glue, and Kinesis, model development using SageMaker, deep learning with EC2 and specialized instances, and production deployment strategies. Key topics covered include exploratory data analysis, feature engineering techniques, algorithm selection, hyperparameter tuning, model evaluation metrics, A/B testing, and monitoring deployed models. Students learn to use AWS SageMaker for end-to-end ML workflows, implement batch and real-time inference, handle model versioning, and ensure security and compliance for ML workloads. The certification covers specialized AWS ML services including Comprehend for NLP, Rekognition for computer vision, Textract for document analysis, and Personalize for recommendation systems. Real-world scenarios include building recommendation engines, fraud detection systems, predictive maintenance solutions, and conversational AI applications. Duration: Self-paced preparation (typically 2-4 months), 3-hour exam. Prerequisites: 1-2 years of ML experience and familiarity with AWS services. Official AWS certification recognized globally by employers.
Generative ai with large language models
0.0
Generative AI with Large Language Models is a comprehensive course that teaches how to deploy and use generative AI models in real-world applications. This course covers the complete lifecycle of large language model development including pre-training, fine-tuning, and deployment strategies. Students learn about transformer architectures, attention mechanisms, and the latest advances in generative AI including GPT, BERT, T5, and other state-of-the-art models. The curriculum covers prompt engineering techniques, retrieval augmented generation (RAG), fine-tuning methods (full fine-tuning, parameter-efficient fine-tuning, LoRA), model alignment through reinforcement learning from human feedback (RLHF), and deployment considerations for production systems. Key topics include understanding scaling laws, computational requirements, model compression techniques, and ethical considerations in generative AI deployment. Students gain hands-on experience with Hugging Face Transformers, building custom training pipelines, and implementing various fine-tuning strategies. The course emphasizes practical applications including text generation, summarization, translation, question answering, code generation, and conversational AI systems. Real-world case studies cover chatbots, content creation tools, and enterprise AI applications. Duration: 3-week intensive course with 4-6 hours per week. Prerequisites: machine learning fundamentals and Python programming experience. Includes hands-on labs with cloud-based GPU resources.
Kaggle learn machine learning micro-courses
0.0
Kaggle Learn offers a comprehensive collection of free, practical machine learning micro-courses designed for hands-on learning. These bite-sized courses cover essential ML and data science skills through interactive coding exercises and real datasets. The curriculum includes Python programming fundamentals, pandas for data manipulation, data visualization, intro to machine learning, intermediate machine learning, feature engineering, data cleaning, and advanced topics like deep learning and computer vision. Each micro-course is designed to be completed in a few hours with immediate practical application. Students work with real datasets from Kaggle competitions and learn industry best practices from experienced practitioners. Key topics covered include handling missing values, categorical encoding, cross-validation, overfitting prevention, ensemble methods, neural networks, and model interpretation. The courses emphasize practical skills over theory, making them perfect for beginners wanting to quickly gain hands-on experience. All courses include exercises, projects, and certificates of completion. Special focus on practical techniques used in Kaggle competitions and real-world data science projects. Duration: 15+ micro-courses, 2-6 hours each, self-paced. Prerequisites: basic programming knowledge helpful but not required. Completely free with practical exercises and datasets.
Cs234: reinforcement learning (stanford)
0.0
CS234: Reinforcement Learning is a comprehensive graduate-level course that covers the foundations of reinforcement learning and decision making under uncertainty. Taught at Stanford University, this course provides both theoretical foundations and practical applications of RL algorithms. The curriculum covers Markov Decision Processes (MDPs), value iteration, policy iteration, Q-learning, deep reinforcement learning with neural networks, policy gradient methods, and advanced topics like multi-agent RL and imitation learning. Students learn to implement core RL algorithms from scratch and understand when to apply different methods. Key topics include exploration vs exploitation, function approximation, deep Q-networks (DQN), actor-critic methods, trust region policy optimization (TRPO), proximal policy optimization (PPO), and model-based RL. The course emphasizes both mathematical rigor and practical implementation with programming assignments in Python. Real-world applications covered include game playing, robotics, recommendation systems, and autonomous systems. Students complete hands-on projects implementing RL algorithms for various domains including Atari games, continuous control tasks, and custom environments. Duration: 10-week quarter course with 3 hours of lectures per week plus assignments. Prerequisites: solid background in probability, statistics, linear algebra, and machine learning. All lecture videos, slides, and assignments available free online.
Mlops specialization by deeplearning.ai
0.0
The MLOps Specialization by DeepLearning.AI teaches how to design an ML production system end-to-end, from data collection to model deployment and monitoring. This 4-course specialization covers the complete machine learning operations lifecycle including data pipeline automation, model versioning, continuous integration/deployment for ML, and production monitoring. Students learn to implement MLOps best practices using tools like TensorFlow Extended (TFX), Kubeflow, Docker, Kubernetes, and cloud platforms. The curriculum includes data validation and versioning, feature engineering at scale, model training automation, A/B testing for ML models, and production deployment strategies. Key topics covered include data drift detection, model performance monitoring, automated retraining pipelines, infrastructure as code, experiment tracking, and governance frameworks. Students gain hands-on experience with real-world MLOps tools and platforms including Google Cloud AI Platform, AWS SageMaker, and open-source solutions. The specialization emphasizes practical skills for maintaining ML systems in production including debugging production issues, scaling inference services, and ensuring model reliability. Duration: 4 courses, approximately 3 months at 6-8 hours per week. Prerequisites: experience with machine learning, Python programming, and basic knowledge of cloud platforms. Includes hands-on projects with industry-standard MLOps tools.
Tensorflow developer certificate program
0.0
The TensorFlow Developer Certificate Program is designed to test a developers foundational knowledge of integrating machine learning into tools and applications using TensorFlow. This hands-on certification program covers building and training neural network models using TensorFlow and Keras, working with real-world image data and natural language processing problems, and understanding how to deploy models on mobile and web platforms. The curriculum includes neural network fundamentals, computer vision with CNNs, natural language processing with RNNs and transformers, time series analysis and forecasting, and model deployment strategies. Students learn to preprocess data, build various neural network architectures, implement regularization techniques, and optimize models for production. Key topics include data augmentation, transfer learning, embedding layers, sequence modeling, attention mechanisms, and mobile/edge deployment with TensorFlow Lite. The certification exam tests practical skills including implementing CNN architectures for image classification, building NLP models for text classification and generation, handling time series data, and deploying models. Duration: Self-paced preparation (typically 1-3 months), 5-hour exam. Prerequisites: basic Python programming and fundamental machine learning knowledge. Official Google certification recognized by employers worldwide.
The Deep Learning Specialization by Andrew Ng is a comprehensive 5-course program that teaches the foundations of deep learning and how to build neural networks. Created by Andrew Ng and deeplearning.ai, this specialization covers neural networks and deep learning, improving deep neural networks through hyperparameter tuning and optimization, structuring machine learning projects, convolutional neural networks for computer vision, and sequence models for NLP. Students learn to build and apply deep learning algorithms, understand the key parameters in neural networks, and implement efficient (vectorized) neural networks. The curriculum includes hands-on programming assignments using Python, NumPy, TensorFlow, and Keras. Key topics covered include forward propagation and backpropagation, different optimization algorithms (Adam, RMSprop), batch normalization, dropout regularization, CNN architectures (LeNet, AlexNet, VGG, ResNet), object detection (YOLO, R-CNN), face recognition, neural style transfer, RNNs, LSTMs, GRUs, word embeddings, and attention mechanisms. Real-world case studies include autonomous driving, sign language recognition, music generation, chatbots, and machine translation. Duration: 5 courses, approximately 3-4 months at 5-7 hours per week. Prerequisites: basic knowledge of programming and machine learning (Andrew Ng ML course recommended). Includes programming assignments, quizzes, and hands-on projects.
Stanford CS231n: Convolutional Neural Networks for Visual Recognition is one of the most comprehensive and rigorous courses on computer vision and deep learning available online. This graduate-level course, taught by renowned Stanford faculty including Fei-Fei Li, Justin Johnson, and Andrej Karpathy, provides deep theoretical understanding and practical implementation of convolutional neural networks for computer vision tasks. The curriculum covers image classification, object detection, semantic segmentation, generative models, and advanced architectures including ResNet, Inception, and attention mechanisms. Students gain hands-on experience implementing CNN architectures from scratch, understanding backpropagation in detail, and working with popular frameworks including TensorFlow and PyTorch. The course emphasizes both mathematical foundations and practical applications with assignments involving real-world datasets and state-of-the-art techniques. Key topics include convolution operations, pooling layers, optimization algorithms, regularization techniques, transfer learning, object detection frameworks (R-CNN, YOLO), generative adversarial networks, and recent advances in computer vision research. Students complete three comprehensive assignments including image classification, neural network training, and a final project. All lecture videos, slides, and assignments are available free online. Duration: 10-week course with 2 hours of lectures per week plus extensive assignments. Prerequisites: proficiency in Python, linear algebra, calculus, and probability.
NVIDIA Deep Learning Institute Fundamentals of Deep Learning is a comprehensive hands-on course that teaches the fundamentals of deep learning through practical experience with real datasets and industry-standard tools. This instructor-led workshop covers neural network architectures, training techniques, and deployment strategies using NVIDIA GPU-accelerated computing. The curriculum includes deep learning fundamentals, convolutional neural networks for computer vision, recurrent neural networks for sequence data, and advanced topics including transfer learning and multi-GPU training. Students gain practical experience with popular frameworks including TensorFlow, Keras, and PyTorch on NVIDIA GPUs. The course emphasizes hands-on learning with real-world datasets including image classification, object detection, natural language processing, and time series analysis. Key learning outcomes include ability to implement CNN and RNN architectures, understand training optimization techniques, apply transfer learning for faster development, and deploy models for production inference. The workshop includes best practices for data preparation, model evaluation, hyperparameter tuning, and performance optimization on GPU hardware. Duration: 8-hour intensive workshop (full day) with optional follow-up sessions. Prerequisites: Basic programming experience in Python. Cost varies by format (self-paced online version available). Includes NVIDIA DLI certificate upon completion and access to GPU cloud instances during training.
Something missing?