Matrix Math Flashcards Preview

Deep Learning > Matrix Math > Flashcards

Flashcards in Matrix Math Deck (41):
1

A single value is known as

scalar

2

scalar

a value with 0 dimensions

3

Lists of values are known as

vectors

4

Types of vectors

row and column

5

Dimensions of vectors

just one: length

6

What is a matrix

a 2 dimensional grid of values

7

What is a tensor?

Any n-dimensional collection of values

8

Locations in matrices are known as

indices

9

Indices nomenclature

Like a 1 layer nested array

10

Numpy is

a C library in python.
Does lots of math operations in Python and is designed to work with matrices.

11

Normal convention for naming numpy

import numpy as np

12

Most common way to with number in NumPy is through

ndarray objects

13

ndarray objects are

similar to Python lists, but can have any number of dimensions
Does fast math operations

14

To declare an ndarray

x = nd.array(5)

15

To get shape of ndarray

nd.shape

16

To reshape an nd array like one that is (4,)

(4,).reshape(1,4)

17

Why do some people use
x = v[:, None]

Adds extra dimension

18

Elementwise operations

Like iterating through and running an operation

19

Requirements for adding two matrices

Have to be the same shape

20

When describing the shape of a matrix how does one describe it?

rows x columns

21

You can only safely run a transpose to multiply if

The data is arranged as rows

22

To get the min, max, mean of a matrix

np.min(array)
np.max(array)
mp.mean(array)

23

How to calculate error in a logistic regression?

It the number of errors

24

What method does one use to minimize the error?

Gradient descent

25

Basic parts of a neural network

Input data, processing, output

26

Individual nodes are called

perceptrons

27

What are weights?

A higher weight means the neural network considers that input more important than other inputs, and lower weight means that the data is considered less important.

28

W vs w

W when it represents a matrix of weights or a w when it represents an individual weight

29

How is an output signal determined?

feeding the linear combination into an activation function

30

What are two ways to go from an AND perceptron to an OR perceptron?

Increase the weights
Decrease the magnitude of the bias

31

AND perceptron

Both must be true to accept

32

OR perceptron

One must be true

33

NOT perceptron

A specific one must be true

34

XOR perceptron

outputs 0 if the inputs are the same and 1 if the inputs are different

35

Gradient is

term for rate of change or slope

36

To calculate rate of change

derivative of a function f(x) gives you another function f​'(x) that returns the slope of f(x) at point x

37

Local minima

where the error is low, but not the lowest

38

SSE is

measure of networks performance. Low means good predictions.

39

np.dot is the same as

Multiplying two matrices and then getting the sum

40

sigmoid(x)

1/(1+np.exp(-x))

41

sigmoid_prime(x)

sigmoid(x) * (1 - sigmoid(x))