Python Experts
What is the functon used to find the length of a list/array?
len(my_list)
What are images for computers? One sentence plz, don't overthink it.
What are numbers?
Define epoch in the context of neural network
What is one complete pass through the entire training dataset?
We had a US president showing up on yesterday's slide, who is that guy?
Who is Abraham Lincoln?
Spell the name of our head instructor out loud in 5 seconds!
What is Millard McElwee?
L = []
print(L[0])
What is going to happen?
Error
Describe the shape of an image, what data structure is used to store an image?
What is a Matrix?
What is the machine learning library we used for Neural Network in Codio?
What is Keras?
Which model is ChatGPT based on?
What is Transformer Model?
Without the softmax function, which of the following characters is going to perish first: Stalin, the sheep, or the US?
for i in range(6){
print(i)
}
Catch syntax error and change it to the correct one.
for i in range(6):
print(i)
Describe Greyscale and RGB channels, what are their shared characteristics, and what are their differences?
Shared Features: Both greyscale and RGB channels have color pixel ranging from 0 to 255.
Differences: 1 channel v.s. 3 channels
Name the 3 major layers for a neural network
What is Input layer, hidden layer, output layer?
In yesterday's class, Conny drew a character on the board for convolution demonstration? What was it?
What is X?
How many numbers of homework you guys need to submit for this week's deep learning class on Canvas?
What is 2?
Name 3 important libraries that we have covered in this class. Please give out their full name and their corresponding alias. Hint: 2 for mathematical operation, 1 for plotting and the plotting one is a bit tricky.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
What is the function used to display an image?
What is plt.imshow(image)?
Write out ReLU function.
What is Max(0, x)?
What is a feature map and what does it represent?
It is the output of a convolution operation after applying a filter to the image.
It measures how well the filter captures the feature of the image. Usually a bigger number represents a higher resemblance
What will happen if in CNN, we use a filter of size 1 * 1?
The convolution steps become meaningless. As applying a 1 * 1 filter is conceptually just a vanilla neural network where each node in the hidden layer gets to see every single input layer nodes
What is the dataset name for the handwritten digits that we covered on Wednesday?
What is MNIST?
Define overfitting!!!
When the model performs well on training data but poorly on new data
What is the primary function/purpose of MaxPooling in CNN?
What is primarily used for "Reduce Computational Load" or "Reducing Image Size"
What are the two human senses that are still challenging for computers to comprehend?
What is Taste and Smell ?