Deep learning differs from traditional machine learning in that it uses multi-layered neural networks to automatically learn features from data, while traditional methods rely on manual feature extraction and simpler algorithms.

Deep learning is only applicable to structured data, while traditional methods handle unstructured data better.

Deep learning requires more manual feature extraction than traditional machine learning.

Traditional machine learning uses deep neural networks for feature learning.

Neural networks are a type of machine learning model that simulate the way human brains operate to recognize patterns and make decisions.

Neural networks are algorithms that only work with structured data.

Neural networks are a form of traditional programming that follows strict rules.

Neural networks are a type of hardware used for data storage.

Activation functions enable neural networks to learn non-linear relationships.

Activation functions are responsible for data preprocessing before training.

Activation functions are used to initialize weights in neural networks.

Activation functions help in reducing the size of the neural network.

Overfitting occurs when a model is too simple and cannot capture the underlying patterns in the data.

Overfitting is when a model performs equally well on both training and unseen data.

Overfitting in deep learning is when a model performs well on training data but poorly on unseen data due to excessive learning of noise and details.

Overfitting happens when a model is trained on too little data, leading to poor generalization.

A convolutional neural network (CNN) is a shallow learning model that processes unstructured data.

A convolutional neural network (CNN) is a type of recurrent neural network used for time series analysis.

A convolutional neural network (CNN) is a deep learning model designed for processing grid-like data, primarily used in image analysis.

A convolutional neural network (CNN) is primarily used for natural language processing tasks.

RNNs are designed to forget all previous inputs after processing each one.

RNNs work by maintaining a hidden state that updates with each input in a sequence, allowing them to learn from previous inputs.

RNNs process inputs in parallel without any hidden state.

RNNs only work with fixed-size input data and do not handle sequences.

To increase the number of neurons in the network.

To enhance the learning rate during training.

To ensure all neurons are activated at all times.

The purpose of dropout in deep learning is to prevent overfitting by randomly deactivating a subset of neurons during training.

Linear Regression

Stochastic Gradient Descent (SGD), Adam, RMSprop, Adagrad, Adadelta, Nesterov Accelerated Gradient (NAG)

K-Means Clustering

Principal Component Analysis

A process where models are only used for their original tasks.

Transfer learning is a technique in machine learning where a model trained on one task is adapted for a different but related task.

A technique that requires training a model from scratch for every task.

A method to improve the accuracy of a single task model.

Test datasets are used to create new models.

Training datasets are used for data storage.

Training datasets help the model learn, validation datasets help tune the model, and test datasets evaluate the model's performance.

Validation datasets are for data visualization.

To increase the number of neurons in the network.

To randomly initialize the weights of the network.

The purpose of backpropagation in neural networks is to optimize the weights by minimizing the error through gradient descent.

To collect data for training the model.

Use only training loss to assess performance.

Use metrics like accuracy, precision, recall, F1 score, and AUC, along with validation and cross-validation.

Evaluate based solely on the number of layers in the model.

Ignore validation data and focus on training data only.

Pandas

Scikit-learn

NumPy

TensorFlow, PyTorch, Keras, MXNet

Fairness, transparency, privacy, accountability, and societal impact.

Cost-effectiveness

Data visualization techniques

Algorithm complexity