Courses from 1000+ universities
AI got cheap enough that Duolingo’s most expensive plan may not survive it. I read the earnings call transcript and opened the app to see what is actually changing for learners.
600 Free Google Certifications
Artificial Intelligence
Data Science
Cybersecurity
L'Italiano nel mondo
Introduction to HTML5
Umano Digitale
Organize and share your learning with Class Central Lists.
View our Lists Showcase
Master object detection fundamentals and implement YOLO architectures from scratch using PyTorch, covering key metrics like IoU, NMS, and mAP with hands-on coding.
Discover how to implement popular machine learning algorithms from scratch using Python, covering KNN, regression, clustering, SVM, and neural networks.
A walkthrough of Whisper’s robust multilingual speech recognition model, covering weakly supervised scaling, architecture, evaluation, and long-form transcription.
An explanation of how GCNs propagate node representations through graph connections, linking message-passing intuition to spectral derivation and node classification.
A theoretical and practical walkthrough of topic modeling with LDA, focusing on posterior inference and collapsed Gibbs sampling.
Build a variational autoencoder from scratch in PyTorch, train it on MNIST, and use it for inference and generation.
Learn to create AI art in Midjourney by iterating prompts, refining keywords, and using parameters to guide generated images.
Implement an SRGAN in PyTorch from scratch, building its generator, discriminator, residual blocks, and VGG-based loss for image super-resolution.
A simple deep learning solution for a Kaggle facial keypoint detection competition, predicting 15 facial landmark coordinates from small grayscale images.
Build a high-performing diabetic retinopathy image classifier by iteratively improving a baseline with preprocessing, augmentation, paired-eye information, and higher resolution.
Implement ProGAN from scratch, covering progressive growing, fade-in layers, training setup, and evaluation for high-resolution image generation.
A paper walkthrough of ProGAN’s progressive training strategy and key techniques for stabilizing high-resolution image generation.
Implement CycleGAN from scratch for unpaired image-to-image translation between domains such as horses and zebras.
Implement EfficientNet from scratch in PyTorch, including squeeze-and-excitation, inverted residual blocks, stochastic depth, and model scaling.
Solve Santander's anonymous transaction prediction Kaggle challenge with a simple neural network, feature engineering, and a top-1% solution without ensembling.
Get personalized course recommendations, track subjects and courses with reminders, and more.