Regression and Text Classification tasks with FCNs Flashcards
What is the key difference between classification and regression?
Classification predicts categorical labels, while regression predicts continuous numerical values.
What does MSE measure in regression tasks?
MSE measures how badly the AI system’s predicted values deviate from the actual values.
What is the function of an output layer in a neural network?
It transforms hidden layer outputs into final predictions using activation functions.
What is the purpose of the Boston Housing Price dataset?
It is used to predict house prices based on various features using regression models.
What are two common data normalization techniques?
Min-Max Scaling and Z-Score normalization.
How many hidden layers and units are used in the Boston Housing model?
Two hidden layers, each with 64 units.
What is the IMDB dataset used for?
It is used for binary classification of movie reviews into positive and negative sentiments.
Why is one-hot encoding necessary for text classification?
Neural networks cannot process varying-length text, so one-hot encoding converts text into fixed-length binary vectors.
What activation function is used in the last layer of the IMDB classification model?
Sigmoid activation function.
What loss function is used for binary classification tasks?
Binary cross-entropy.
How many categories does the Reuters dataset have?
46 different topics.
What loss function is used for multiclass classification in the Reuters dataset?
Categorical cross-entropy.
Which optimizer is commonly used in both regression and classification models in the slides?
Adam optimizer.
How is Min-Max Scaling applied to data?
X(scaled) = (X - X(min))/(X(max) - X(min))
What is the formula for Z-Score normalization?
X(standard) = (X - mean)/variance
