Ans Imran

The said jupyter notebook was updated on June 27. Somewhere around 11 June, 2023, I trained a small Convolutional Network. The steps were following: Defined the Convolutional Base - used conv2d & maxpooling layers Defined the Classifier Base - used dense and flatten layers Used the MNIST Data Standardized the data - Dividing each pixel by 255 Performed one-hot-encoding of  test and train data labels Defined the compiler Trained the Model The accuracy on training data was more than 99 %. Made Predictions on test data The accuracy on test data was more than 98 %. Visualized the images You can see the jupyternotebook here ----- A Simple Convolutional Network

 My name is Ans IMRAN.   Currently I am doing M1 in Physics (  Paris Physics Master ) from Sorbonne University and Universite Paris Cite. Sorbonne University is considered to be a very prestigious university in France. In parallel, I am also enrolled in an online program in Artificial Intelligence . This program is in-part funded by Pakistani Government. It is run by  PIAIC ( Presidential Initiative for Artificial Intelligence and Computing ). So, right now, I am studying both Artificial Intelligence and Physics at a somewhat advanced level. After finishing my M2 in Physics and my online program in Artificial Intelligence, I'll decide whether I want to go for Physics or whether I want to go for AI. On this blog I'll be sharing the decisions that I made and the things that I learned along the way! Here is a link to my github repository . Here you can see some of the things that I have already learned. E.g: Python, Numpy, Pandas etc...   Wish me Bonne Chance! Au...