Neural Network Parts
What directly plays into the final prediction?
The output layer/weights.
What is computer vision?
The way machines interpret images.
If Netflix recommends a movie to you because other users with highly similar watch histories loved it, what type of filtering is it using?
Collaborative filtering.
What is the input layer?
The starting point that receives raw data.
Weights are representations.
FALSE BRO.
What part of the neural network receives the raw data (pixels and numerical values) before any processing begins?
The input layer.
What do the earliest hidden layers of a computer vision neural network typically look for in an image?
The edges, lines, and borders.
If Spotify recommends a song to you purely because its genre, tempo, and instruments match songs you actually like, what filtering method is it using?
Content-based filtering.
What are hidden layers?
Middle layers that extract complex features and patterns from the data.
Embeddings are influences.
AGAIN, FALSE.
What are the mathematical values called that increase/decrease the strength of a signal between neurons?
The weights.
How are pixels depicted in computer vision?
Through numerical values.
If responses are biased, then what will happen to the neural network?
It will probably have inaccurate predictions.
What is the output layer?
The final layer that produces the network's actual prediction.
If embeddings are close together, they are modifiers.
TRUEEEEEE
What do the hidden layers do?
They apply the weights to the final prediction.
What does computer vision do when handed a fresh set of data, and a preset of options?
It guesses, and recognizes patterns.
Why is it important for recommendation systems to have a diverse set of data?
So you can find new things to get recommended, instead of getting the same thing again and again (peak song: Again & Again---The Bird and The Bee)
What is a false negative?
An answer that overlooks the right answer, and completely dodges it.
The main job of a hidden layer is to determine how strongly one input influences a calculation.
TRUEEEEEEE.
What are neurons in an neural network?
Processors of information and data.
Machine Learning finds....
patterns by learning from its mistakes.
What problem do recommendation systems face when a brand-new user signs up or a brand-new item is added, leaving the model with no historical data to work with?
The Cold Start Problem.
What is a false positive?
A wrong "right" answer.
Deep learning simply refers to neural networks that contains multiple hidden layers between the input and the output.
TRUEEEE.