5 Data Science Interview Questions Part VI
1. What are tensors?
Tensor is a generalization of an n-dimensional array.
- A scalar is a 0-dimensional tensor.
- A vector is a 1-dimensional tensor.
- A matrix is a 2-dimensional tensor.
If your array has 3 or more dimensions, they are just called tensors.
2. Why do you use tensors in deep learning?
Statistically, tensors allow the representation of higher dimensional relationships in the data, facilitating the identification of hidden relationships.
Computationally, they allow for faster training and inference.
3. What is dropout?
Dropout is a regularization method that consists of randomly “dropping units from the neural network during training”. It helps minimize overfitting and improves the performance of the model.
4. What are auto-encoders?
Auto-encoder is an unsupervised artificial neural network that learns to encode and compress the data. Furthermore, it learns to reconstruct the data (decode) by generating inputs as close to the input as possible.
Auto-encoders are good for dimensionality reduction and therefore, reducing the computational cost and time.
5. What is the use of an activation function in a deep learning model?
activation functions are used to add non-linearity properties to the network, which helps the model learn more complex relationships and identify complex patterns from the data.