¡Tu no aprendes!
Detecting spam emails using a model trained on labeled data is an example of this type of learning, which uses a validation dataset to test the model's performance.
Supervised Learning
Logistic regression is used for this type of problem, separating output into different categories.
Classification
This algorithm is used to minimize the loss function in a machine learning model. It has the analogy of going down a hill.
Gradient Descent
A convolutional neural network is used primarily for this application because it extracts features from the red, green, and blue channels.
Image Process
In Natural Language Processing, vocabulary is tokenized into these entities that have magnitude and direction.
Covered in day 1's research, these two plants are actually in the same family.
Potato and Tomato.
This model is used to predict continuous values based on input features. It has a guaranteed-to-be-reachable global minimum under gradient descent.
Linear Regression
This hyperparameter determines the "step size" in gradient descent. Setting this hyperparameter too high risks making the model "bounce around."
Learning Rate
Name 2 of the 5 Convolutional Neural Networks you presented on Thursday.
AlexNet, LeNet, ResNet, VGG, Inception
This is a common issue when using a Recurrent Neural Network for sequence prediction tasks.
They can easily forget earlier information in the sequence.
This learning method involves finding patterns and structures in data without labeled responses.
Unsupervised Learning
This activation function returns 0 if the input is negative; otherwise, it returns the input.
ReLu
This entity grants machine learning models to predict non-linear relationships. Sigmoid is this entity.
Activation Function
This part of the Convolutional Neural Network applies a series of filters to detect features in image.
Convolutional Layers
This is the name of the cells in a Recurrent Neural Network that help it remember long-term dependencies
Memory Cell
This learning method requires models to make decisions through trial and error, eventually maximizing long-term gain.
Reinforcement Learning
Draw the ReLU function
-----///////
In deep learning, the weight initialization process uses this method to set the neural network's initial weights.
Assign Weights Randomly
This technique reduces the spatial dimensions of the feature map. Its most common variance is "max."
Pooling
This type of RNN architecture is designed to better handle long-term dependencies.
Long Short-Term Memory
The term "Artificial Intelligence" was coined at this higher education institution.
Dartmouth College
This is the most common loss function for logistic regression.
Cross-Entropy
In gradient descent, this term refers to a gradient that is too small to contribute to backpropagation.
Vanishing Gradient
This activation function is used in most Convolutional Neural Networks
ReLU
Outside of long-term dependency, Long Short-Term Memory is also designed to address this problem, preventing earlier layers from being "not important."
Vanishing Gradient