Draw a program flow using symbols

Create a program that is directly executable by a computer

Explain program logic using simple language

Replace high-level programming languages

Create a user interface

Iteration

Sequencing

Decision

Branching

Debugging

Branching

Iteration

Sequential

Modular

Recursion

Output

Input

Process

Decision

Terminal

A program that computes the area of a rectangle

A loop with no stopping condition

A program that prints “Hello World” once

A simple branching program

A data-sorting program

Binary Search

Sequential Search

Bubble Sort

Quick Sort

Merge Sort

Problem analysis

Creating a flowchart/pseudocode

Program implementation

Program testing

Documentation

Input

Output

Process

Start/End

Decision

Make a program run directly

Illustrate the program’s logic flow

Produce calculation results

Test program errors

Create data tables

Input/Output

Process

Decision

Storage

Documentation

More concise

Faster to execute

Easier to understand visually

Closer to machine language

Rectangle

Diamond

Oval

Arrow

Parallelogram

Rectangle

Arrow line

Oval

Triangle

Box with dashed border

Requirements analysis

Coding

Documentation

Debugging

Program logic design

The program does not run

The algorithm is not finite

The program is hard to understand

The flowchart cannot be printed

The flowchart is too long

Turn right

Move forward one step

Turn left

Move backward one step

Jump to the next block

moveForward();

turnLeft();

turnRight();

stepLeft();

rotate();

5 times moveForward();

repeat 5 { moveForward(); }

while (true) { moveForward(); }

if (pathAhead) moveForward();

turnLeft(); moveForward();

Repeat steps forever

Move forward if there is a path ahead

Turn left at a dead end

Choose the right path

Stop the program

Move once if there is a path

Keep moving as long as there is a path ahead

Stop when encountering a turn

Choose the right way

Turn left at an intersection

The number of neural network layers

The total amount of training data

One full pass through the entire training data

The number of neurons in the hidden layer

The activation function used

$$f(x) = \frac{1}{1+e^{-x}}$$

$$f(x) = \max(0, x)$$

$$f(x) = \tanh(x)$$

$$f(x) = \log(x)$$

$$f(x) = e^x$$

Speed up computation

Test the model’s generalization ability

Reduce the amount of data

Eliminate errors

Make the model more complex

Reduce the training data

Add data or use regularization

Stop training earlier

Remove hidden layers

Increase the learning rate without limit