Define Pointer

1

int* ptr;

This declares ptr as a pointer to int

Operators

• * Operator
  1. Means “Pointer” when assignment and initialization.
  2. Means “Go To” the address it stored.
• & Operator
  Means “Get Address”.

Usage:

1. Initialization

1
2

int i=10;
int* ptr = &i;

2. Assignment

1
2
3

int i = 10;
int *ptr;
ptr = &i;

List Pointer

Let’s take this as an example.
Notice that type of tha array must match that of the pointer.

1
2

int a[10];
int *ptr;

Assign the address of a[0] to pointer ptr.

1

ptr = &a[0];

In C, this could be replaced by:

1

ptr = a;

Also, because elements in an array are continuous:

1

*(ptr+i) == a[i] == *(a+1) // True

Remind that a could not change, but ptr could change.

String Pointer

Malloc Function.

Basic

To begin, load the program:

1

> gbd [program name]

or have core dumped:

1

> gdb [program name] [core file]

Then set breakpoints(as follow) and run the program:

1

> run [parameters]

Breakpoints

Set Breakpoint

Two ways to set break points

• Set break point by function name:

1

> break [function name]

• Set break point by line number:

  1	> break [line number]

Show Breakpoints

1

> ib

or

1

> info breakpoint

Delete breakpoint

To delete all break points:

1

> delete

Debug in Breakpoint

Print

Usage:

1

> print([variable])

Next:

Usage:

1

> next

or

1

> n

Linear regression could not yield proper result in classification Problem, because outliers could affect the result.

Logistic Regression is still a classification algorithm, even though it is called regression.

Hypothesis Function

Logistic regression uses the “Sigmoid Function” (or “Logistic Function”) as hypothesis function:

will give us the probability that our output is 1.

Decision Boundary

Decision boundary is on , also
It is a feature of the trained model (parameters).

Non linear decision boundaries:
Just create new features to fit.

Cost Function

$
J(\theta) = \frac{1}{m} \sum_{i=1}^m \mathrm{Cost}(h_\theta(x^{(i)}),y^{(i)})

$

Vectorized version:

Why not Using Squared Error？

If using squared error on sigmoid function, it will be non-convex, so that we can’t use gradient descent.

Optimization

Derivative for Gradient Descent

It looks identical to linear regression
$
\frac{d}{d\theta_j}J(\theta)=\frac{1}{m}\displaystyle\sum_{i=1}^m{(h_\theta(x^{(i)})-y^{(i)})x^{(i)}_j}

$

$
J(\theta_0,\theta_1,\theta_2, …) = \frac{1}{m}\displaystyle\sum_{i=1}^m{Cost(h_\theta(x^{(i)}), y^{(i)})}+\frac{\lambda}{2m}\displaystyle\sum_{j=1}^n{\theta^2_j}

$

( is regularization parameter)

Update :
$
\theta_j = \theta_j -\frac{\alpha}{m}\displaystyle\sum_{i=1}^m{(h_\theta(x^{(i)})-y^{(i)})x^{(i)}_j}

$
Vectorized implementation:

Basic writing

Paragraphs

Paragraphs in Markdown are just one or more lines of consecutive text followed by one or more blank lines.

You can create a heading by adding one or more # symbols before your heading text. The number of # you use will determine the size of the heading.

1
2
3
4

# The largest heading (an <h1> tag)
## The second largest heading (an <h2> tag)
…
###### The 6th largest heading (an <h6> tag)

Block-quotes

You can indicate block-quotes with a >

1
2
3

In the words of Abraham Lincoln:

> Pardon my French

Styling text

1
2

*This text will be italic*
**This text will be bold**

Both bold and italic can use either a * or an _ around the text for styling. This allows you to combine both bold and italic if needed.

1

**Everyone _must_ attend the meeting at 5 o'clock today.**

Lists

Unordered lists

You can make an unordered list by preceding list items with either a * or a -.

1
2
3
4
5

* Item
* Item

- Item
- Item

Ordered lists

You can make an ordered list by preceding list items with a number.

1
2

1. Item 1
2. Item 2

Nested lists

You can create nested lists by indenting list items by two spaces.

1
2
3
4
5

1. Item 1
  1. A corollary to the above item.
2. Item 2
  * A corollary that does not need to be ordered.
    * This is indented four spaces.

Code syntax highlighting

In-line formats

Use single back-ticks (`) to format text in a special mono-space format. Everything within the back-ticks appear as-is, with no other special formatting.

1

Here's an idea: why don't we take `SuperiorProject` and turn it into `**Reasonable**Project`.

Multiple lines

You can use triple back-ticks (```) to format text as its own distinct block.

1
2
3
4

``` 
x = 0
x = 2 + 2
```

Syntax highlighting

In fenced block, add an optional language identifier for syntax highlighting.
For example, to syntax highlight Ruby code:

1
2
3
4
5

```ruby
require 'redcarpet'
markdown = Redcarpet.new("Hello World!")
puts markdown.to_html
```

Links

Link

You can create an in-line link by wrapping link text in brackets ( [ ] ), and then wrapping the link in parentheses ( ( ) ).

For example, to create a hyper-link to www.github.com, with a link text that says, Visit GitHub!, you’d write this in Markdown: [Visit GitHub!](https://www.github.com).

Image

In-line image:

1

![Alt Msg](URL)

Block image:

1

![Alt Msg][URL]

GitHub Flavored Markdown

underscores in words

Where Markdown transforms underscores (_) into italics, GFM ignores underscores in words. To emphasize a portion of a word, use asterisks (*).

URL auto-linking

GFM will auto-link standard URLs, so if you want to link to a URL (instead of setting link text), you can simply enter the URL and it will be turned into a link to that URL.

1

http://example.com

becomes: http://example.com

Strike-through

GFM adds syntax to create strike-through text, which is missing from standard Markdown.

1

~~Mistaken text.~~

becomes: Mistaken text.

Tables

You can create tables by assembling a list of words and dividing them with hyphens - (for the first row), and then separating each column with a pipe |:

1
2
3
4

First Header  | Second Header
------------- | -------------
Content Cell  | Content Cell
Content Cell  | Content Cell

You can also add extra pipes on the ends:

1
2
3
4

| First Header  | Second Header |
| ------------- | ------------- |
| Content Cell  | Content Cell  |
| Content Cell  | Content Cell  |

Dashes don’t need to match the length of text:

1
2
3
4

| Name | Description          |
| ------------- | ----------- |
| Help      | Display the help window.|
| Close     | Closes a window     |

Finally, by including colons : within the header row, you can define text to be left-aligned, right-aligned, or center-aligned:

1
2
3
4
5

| Left-Aligned  | Center Aligned  | Right Aligned |
| :------------ |:---------------:| -----:|
| col 3 is      | some wordy text | $1600 |
| col 2 is      | centered        |   $12 |
| zebra stripes | are neat        |    $1 |

A colon on the left-most side indicates a left-aligned column; a colon on the right-most side indicates a right-aligned column; a colon on both sides indicates a center-aligned column.

Basic

A matrix has 3 rows and 2 columns, so it is a 3x2 matrix.

• refers to the element in the th row and th column of matrix A
• refers to the element in the th row of the vector.
• In general, all our vectors and matrices will be 1-indexed.
• Matrices are usually denoted by uppercase names while vectors are lowercase.
• “Scalar” means that an object is a single value, not a vector or matrix.
• refers to the set of scalar real numbers.
• refers to the set of n-dimensioned vectors of real numbers

Addition

Matrix dimensions must agree
Plussing is just add each coordinate numbers together.(Minus as well)

Multiplication

Matrix * number

Plussing is just add each coordinate numbers together.(Minus as well)

Matrix * vector

$
([mn][n1] = [m1])
$

Matrix * matrix

Warnings

Two properties of matrix multiple:

• Not commutative.
• Associative.

Inverse & transpose

Inverse

The inverse of a matrix is denoted . Multiplying by the inverse results in the identity matrix.
A non square matrix does not have an inverse matrix. We can compute inverses of matrices in octave with the pinv(A) function.

Invertibility:

Transpose

In other words:

Clone

Branch

List branches

1
2
3

> git branch -r // remote branches

> git branch -a // all branches

Commit

1

> git commit

Merge

Pull

1

> git pull <远程主机名> <远程分支名>:<本地分支名>

Push

1

> git push <远程主机名> <本地分支名>:<远程分支名>

Fetch

1

> git fetch <远程主机名> <分支名>

Checkout

1

> git checkout

Control Flow

for loop

1
2
3

for x= [initVal:step:endVal or valArray]
    % body
end

Basic Calculation

Inequity Note, not “!=”

1 ~= 2  % true

Variable assignment:

a = 3; % semicolon suppresses output

Displaying them:

1
2
3
4
5
6
7
8

a = pi
disp(a)
disp(sprintf('2 decimals: %0.2f', a))
disp(sprintf('6 decimals: %0.6f', a))
format long
a
format short
a

Matrix Calculation

1. To create a matrix:

1
2
3
4
5

A = [1 2; 3 4; 5 6]

B = [1 2;
	 3 4;
	 5 6]

Or using a built-in function

1
2
3
4
5

ones(row, col)  % same as C = [2 2 2; 2 2 2]
zeros(row, col)
rand(row, col)  % from a uniform distribution (range [0, 1])
randn(row, col) % from a normal distribution (mean=0, var=1)
eye(row)        % 4x4 identity matrix

2. Get informations from matrix

1
2
3
4

size(M)    % return a 1x2 matrix: [(number of rows) (number of columns)]
size(M,1)  % number of rows
size(M,2)  % number of cols
length(M)  % size of longest dimension

3. Manipulating matrix

1
2
3
4
5
6
7
8
9
10
11
12
13

%% indexing
A(3,2)  % indexing is (row,col)
A(2,:)  % ":" means every element along that dimension
A([1 3],:) % print all  the elements of rows 1 and 3

A(:,2) = [10; 11; 12]     % change second column
A = [A, [100; 101; 102]]; % append column vec
A(:) % Select all elements as a column vector.

% Putting data together 
% (A & B are matrics posessing same dims)
C = [A B] or [A,B]        % concatenating A and B matrices side by side
C = [A; B]                % Concatenating A and B top and bottom

Plotting

1
2
3

w = -6 + sqrt(10)*(randn(1,10000))  % (mean = -6, var = 10)
hist(w)     % plot histogram using 10 bins (default)
hist(w,50)  % plot histogram using 50 bins

Basic

Head Part

1
2

#include <stdio.h> // from system
#include "xx.h" // from current folder

Loops

For loops

Usage:

1
2
3
4

for (init; condition; increment)
{
   ...
}

Tips:

1. In “init” part, remember to create a new variable.
2. Start with 0, use i<n to loop n times.

While loops and do-while loops

Usage:

1
2
3
4
5
6
7
8
9

while(condition)
{
   ...
}

do
{
   ...
} while (condition);

Array

To create an array

1

int a[4];

Initialization

1. You could initialize the elements partly: when elements provided is less than those in the array, only former elements were assigned.

1

int a[10]= {0, 1, 2};

2. You could only assign elements one by one even assigning same value:

1

int a[10]= {1, 1, 1, 1, 1, 1, 1, 1, 1, 1};

3. Not providing the length is allowed:

1
2
3

// these two are equal
int a[]={1,2,3,4,5};
int a[5]={1,2,3,4,5};

Pointer

Click for more about pointer in c

Struct

Click for more about struct in c

Linear regression is used when predicting numeric output from numeric input that has linear relationship.

Hypothesis Function

For convenience, define (to use matrix for calculation).

Cost Function

Squared Error

$
J(\theta) = \frac{1}{m}\displaystyle\sum_{i=1}^m{Cost(h_\theta(x^{(i)}), y^{(i)})}

$

$
Cost(h_\theta(x), y)=\frac{1}{2}(h_\theta(x)-y)^2

$

Optimization

Derivative for Gradient Descent

$
\frac{d}{d\theta_j}J(\theta)=\frac{1}{m}\displaystyle\sum_{i=1}^m{(h_\theta(x^{(i)})-y^{(i)})x^{(i)}_j}

$

Update :
$
\theta_j = \theta_j -\frac{\alpha}{m}\displaystyle\sum_{i=1}^m{(h_\theta(x^{(i)})-y^{(i)})x^{(i)}_j}

$

Vectorized implementation:

Normal equation

See Normal Equation for Linear Regression