Logistic Regression (Basic stuff).1. Probability calculation using sigmoid function .2. Log loss (since output values change at inconsistent rate).3. Regularization (L2 & L1)

Day 9/150.🦀 Rust – 2h.• Enum methods & match refinements.• Option & Result usage patterns. 📈 DSA – 3h.• Solved 3 Mediums on Leetcode.• Reviewed time complexities. 🧠 AI/ML – 2h.• Regularization (L1 vs L2).• Visualizing model complexity vs error. 📚 Academics (Day 13) –

Built a Neural Network from scratch using NumPy.- Trained on MNIST for digit classification.- Used a 3-layer setup with ReLU & Softmax.- Tested optimizers: SGD, Momentum, Adam.- Best result:~97.6% val accuracy with SGD.- Applied Dropout and L2 regularization to reduce overfitting

Support Vector Machine(SVM).1. SVM math intuition .2. SVM kernel & it’s types (low dimension to high dimension plotting of vectors).3. Regularization of plane(decision boundary)

Graph is a Natural Regularization: Revisiting Vector Quantization for Graph Representation Learning.

Deep Learning Update . >Completed Perceptron and explored its limitations .>Built a basic neural network from scratch.>Implemented MLP with forward and backward propagation.>Studied Gradient Descent in DL.>Covered VGP and Early Stopping.>Added Dropout layers for regularization

this is such a clean RL objective and it's got the regularization baked in, but you can't really use it without regularization like you can GRPO (where you just delete the KL term). i guess if you update the old policy with each step then that lets it go wherever in policy space

Machine Learning Models. ,. 1. Supervised Learning. Bayesian-based.Tree-based.Instance-based.Regularization-based.Neural Network-based.Ensemble-based. 2. Semi-supervised Learning. Inductive-based.Transductive-based. 3. Unsupervised Learning. Clustering-based.Dimensionality

Three months window for regularization of immigration protocols expires today the first of August 2025. Enforcement and prosecution to follow those who don't comply. #eCERPAC. #RenewedHope .President Tinubu

#Regularization .#SSEsAEOs

Punjab’s govt colleges are being left to die. 1,158 professors who served for years are denied regularization by the Supreme Court — while the Punjab Govt stands silent. This isn’t just about jobs. It’s about the collapse of public higher education. File the review petition.

5. Data Science: Machine Learning. Build a movie recommendation system using popular data science techniques. Learning Objectives: Basics of ML, cross-validation, popular algorithms, recommendation system building, understanding regularization. 🔗

SPOs in J&K risk their lives daily against terrorism but earn just ₹6K–18K/month. Home Guards in other UTs earn ₹30K–35K with safer duties. Appeal to @AmitShah Ji for regularization, salary revision & social security for these unsung heroes. #JusticeForSPOs #JKPolice

J&K Casual Daily Wagers Forum protests in Srinagar demanding job regularization, fair wages. #JammuAndKashmir

Some people don't realize that the interplay between optimization landscapes and implicit bias of gradient descent reveals that converged minima often encode inductive priors more strongly than explicit regularization, challenging traditional notions of model capacity control.

Multi-Task Learning: Jointly optimize action and intent recognition to share features and boost generalization. Mathematical Rigor: Ground models in optimization (gradient descent), regularization (L2, dropout), and information theory (mutual information). Kalman Filtering:

With the sanitary workers’ protest seeking job regularization crossing 10 days, labour body Uzhaipor Urimai Iyakkam has moved the Madras High Court to quash the GCC’s ₹276 crore contract awarding garbage disposal in Zones 5 and 6 to a private firm. The petitioner fears 1,953

Punjab’s govt colleges are being left to die. 1,158 professors who served for years are denied regularization by the Supreme Court — while the Punjab Govt stands silent. This isn’t just about jobs. It’s about the collapse of public higher education. File the review petition.

in 20min you are going to be caught up with the adamw optimizer with this new technical tutorial. why weight decay on adam?.why not L2 regularization?.how come it's 7 LOC?. all of these questions answered and hammered down with the language level that a toddler would nod along