Shubham Panchal
  • Member for 3 years, 1 month
  • Last seen more than a month ago
2 answers
3 votes
141 views
Is an artificial intelligence software or hardware?
Accepted answer
3 votes

AI or Artificial Intelligence is nothing but intelligence but in its artificial form. Intelligence comes from formation of rules and patterns in the data which is seen or on which it is trained. For ...

View answer
1 answers
7 votes
1k views
Which algorithm is used in the robot Sophia to understand and answers the questions?
Accepted answer
2 votes

Sophia ,first , has all the questions and corresponding answers preprogrammed. It is a system which is a hybrid of Bayes Text classification and decision trees. It may consist of a speak recognizer ...

View answer
1 answers
1 votes
611 views
LSTM language model not working
Accepted answer
1 votes

You can follow the below steps : LSTMs are slower in terms of convergence. They take much time to train amd thereby give better results. Try training the network for a more number of epochs like 50 ...

View answer
1 answers
1 votes
70 views
What AI designs are suited for producing title replacements?
1 votes

You can use a seq2seq LSTM model. Giving a article will generate a title for it. You need to train it on data which comprises news articles and their corresponding titles. You can refer here and here....

View answer
1 answers
1 votes
31 views
Training by one batch of examples, what does it mean
1 votes

This would mean that there is only one gradient update on your dataset of 16 samples. If you are taking the mean of all the squared errors ( mse ), then the loss of each sample is contributing to the ...

View answer
1 answers
0 votes
30 views
Getting better results in improving the configuration
Accepted answer
1 votes

Try using an RMSProp optimizer instead of Adam optimizer. Also try decreasing the batch size and keep a small learning rate like 0.001.

View answer
1 answers
-1 votes
103 views
Number of units of the last layer
Accepted answer
1 votes

This is simply because they have different loss functions. Categorical cross entropy measures the logarithmic loss for every neuron in the output layer. For binary crossentropy, we have 1 neuron in ...

View answer
1 answers
1 votes
83 views
Relationship between input range and channel means, standard deviations for CNNs
0 votes

Standardization of the pixel values would bring it mean close to 0 and standard deviation as 1. Normalization will squeeze the values between 0 and 1. For RGB images, divide each channel i.e each ...

View answer
2 answers
2 votes
97 views
Can the same input for a plain neural network be used for a convolutional neural network?
0 votes

Neural networks generally deal with 1 D data. For example, the data for a NN would be ( 10 , 12 ) , where 10 is the number of samples. Convolutional neural networks generally deal with 1D, 2D and 3D ...

View answer
3 answers
9 votes
2k views
Do GANs come under supervised learning or unsupervised learning?
-1 votes

GANs are unsupervised machine learning algorithms. According to Wikipedia, unsupervised algorithms are : Unsupervised learning is a branch of machine learning that learns from test data that has ...

View answer