Categories
Artificial Intelligence (AI) Mastering Development

Finding the ‘ultimate resolution’ of an ANN

I want to use a neural network to predict the refractive index of a solution. My thinking is, instead of immediately training on many samples, I will first find the ‘ultimate resolution’ of the network given the experimental apparatus I am using. What I mean is I will make two different solutions which have refractive […]

Categories
Artificial Intelligence (AI) Mastering Development

Monte Carlo Exploring Starts broke for 2048 game AI

I implemented a MCES for 2048 (the game), with a quality function implemented as a neural net of a single layer. The starts are created with 6 cells filled with values between 64 and 1024, two cells are 1024 an ther other 8 cells are filled with 0. The game is then progressed until the […]

Categories
Artificial Intelligence (AI) Mastering Development

Which paper introduced the term “softmax”?

Nowadays, the softmax function is widely used in deep learning and, specifically, classification with neural networks. However, the origins of this term and function are almost never mentioned anywhere. So, which paper introduced this term?

Categories
Artificial Intelligence (AI) Mastering Development

It can be treted forcasting with neural networks only as a supervised learning (regression) problem?

I have recently made a work about the application of neural networks to time series forecasting and I treated this as a supervised learning (regression) problem. I have came across with suggestion of treated this problem as unsupervised, semi-supervised or reinforcement learning problem. The ones that made this suggestion didn’t know how to explain me […]

Categories
Artificial Intelligence (AI) Mastering Development

How to “forward” updated NN model to a transferred model?

I’ve trained a robot to walk in a straight line for as long as it can (using TD3), and now I’m using that pre-trained model for two new models with separate purposes: 1. Walk to a specific point and halt (adding target position to the NN inputs); 2. Walk straight at a specified velocity (adding […]

Categories
Artificial Intelligence (AI) Mastering Development

rotate output also

I am training a neural network that takes an input (H,W,3) and has output of size (H’,W’,C). Now, to augment my dataset since I only have 45k images I am using the following in my custom data generator def Generator(): img=cv2.imread(trainDir+’\’+imgpath) img=tf.keras.preprocessing.image.random_rotation(img,20m,row_axis=0,col_axis=1,channel_axis=2) output_mask=np.load(trainDir+’\’+maskpath) yield(img/255-.5,output_mask) What I am really wondering is since I am rotating my […]

Categories
Artificial Intelligence (AI) Mastering Development

Max slow to propagate gradient

Isn’t it true that using torch.max(…) over a softmax will be much slower because there is not a smooth gradient? Max basically zeros out the gradients of all the non-maximum values. Especially at the beginning of training, this means it is zeroing out potentially useful features simply because of random weight initialization. Wouldn’t this drastically […]

Categories
Artificial Intelligence (AI) Mastering Development

Neural Network is not learning a very simple task

I am a complete beginner in the area. I implemented my first neural network following the online book "Neural Networks and Deep Learning" by Micheal Nielsen. It works fine with classifying handwritten digits. Achieving ~9500/10000 accuracy on the test data. I am trying to train the network to determine whether $x > 5$ where $x$ […]

Categories
Artificial Intelligence (AI) Mastering Development

which conferences/journals to follow for ML, Deep Learning, Computer Vision, Natural Language Processing?

I have recently completed the Deep learning specialization by deeplearning.ai from Coursera. My instructor has advised me to look at some papers for opinion modeling, aspect extraction from video. Which conferences and journals should I follow for these topics and also neural nets, machine learning, deep learning, computer vision, and natural language processing? Your help […]

Categories
Artificial Intelligence (AI) Mastering Development

Can rewards be decomposed into components?

I’m training a robot to walk to a specific $(x, y)$ point using TD3, and, for simplicity, I have something like reward = distance_x + distance_y + standing_up_straight, and then it adds this reward to the replay buffer. However, I think that it would be more efficient if it can break the reward down by […]