“Learning, building,... • 7h
Day 3/60: "Perceptron ek Jugaad " Good morning my name is Anuj Tongse and, Today is day 3 of understanding Ai Deep learning Today was all about the "Perceptron jugaad(trick)" :-the actual mechanism that allows a model to learn from its mistakes. It's one thing to see a decision boundary; it's another to understand how it moves. What I covered today: Perceptron Tricks & Transformation: Understanding how weights and bias shift the line to classify points correctly. : Labeling Regions: How we mathematically define which side of the line is "Class O" vs. "Class 1" using the step function. The Reality Check :- While the Perceptron is the grandfather of Neural Networks, it's not perfect. It struggles with: Non-linearly separable data:- If the points are mixed, it never stops looping. Lack of convergence: It doesn't tell us how confident it is; it just gives a hard Yes or No. #DeepLearning #ArtificialIntelligence #60DayChallenge #MachineLearning #Neural Networks #Day3




Gen AI, Cybersecurit... • 1y
The AI Legends #75 Days Day 7: Frank Rosenblatt Rosenblatt is best known for his work in developing the perceptron, an early form of artificial neural network. History: Frank Rosenblatt was an American psychologist born on July 11, 1928, in New York
See More
Gen AI, Cybersecurit... • 1y
The AI Legends #75 Days Day 31: Jürgen Schmidhuber Jürgen Schmidhuber is best known for his pioneering work in the development of neural networks, particularly the Long Short-Term Memory (LSTM) networks. History: Jürgen Schmidhuber was born on 1963.
See More
Download the medial app to read full posts, comements and news.