Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Python Application Case
Search
Vicky Vernando Dasta
January 27, 2018
Programming
0
65
Python Application Case
Python Workshop @ Kongkow IT 2018
Vicky Vernando Dasta
January 27, 2018
Tweet
Share
More Decks by Vicky Vernando Dasta
See All by Vicky Vernando Dasta
PYTHON 101
vickydasta
0
110
Other Decks in Programming
See All in Programming
Designing Repeatable Edits: The Architecture of . in Vim
satorunooshie
0
220
CSC305 Lecture 13
javiergs
PRO
0
350
Introducing RemoteCompose: break your UI out of the app sandbox.
camaelon
2
430
Kotlinで実装するCPU/GPU 「協調的」パフォーマンス管理
matuyuhi
0
240
CSC305 Lecture 10
javiergs
PRO
0
330
CSC305 Lecture 11
javiergs
PRO
0
320
SidekiqでAIに商品説明を生成させてみた
akinko_0915
0
110
ドメイン駆動設計のエッセンス
masuda220
PRO
15
7.3k
AsyncSequenceとAsyncStreamのプロポーザルを全部読む!!
s_shimotori
1
220
開発組織の戦略的な役割と 設計スキル向上の効果
masuda220
PRO
10
2k
なんでRustの環境構築してないのにRust製のツールが動くの? / Why Do Rust-Based Tools Run Without a Rust Environment?
ssssota
14
47k
釣り地図SNSにおける有料機能の実装
nokonoko1203
0
200
Featured
See All Featured
How to Ace a Technical Interview
jacobian
280
24k
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
508
140k
Building Flexible Design Systems
yeseniaperezcruz
329
39k
Typedesign – Prime Four
hannesfritz
42
2.8k
Testing 201, or: Great Expectations
jmmastey
46
7.7k
The Illustrated Children's Guide to Kubernetes
chrisshort
51
51k
Bash Introduction
62gerente
615
210k
The Cult of Friendly URLs
andyhume
79
6.7k
4 Signs Your Business is Dying
shpigford
186
22k
The World Runs on Bad Software
bkeepers
PRO
72
11k
Large-scale JavaScript Application Architecture
addyosmani
514
110k
Save Time (by Creating Custom Rails Generators)
garrettdimon
PRO
32
1.7k
Transcript
PYTHON WORKSHOP HAFIZHAN: WEB APPLICATION WITH PYTHON FLASK VICKY: PYTHON
101 AND APPLICATION IN SECURITY & MACHINE LEARNING
Workshop repo github.com/vickydasta/kongkow-python
About • Student @ physics dept. UR • Research assistant
at photonic lab @ UR • Research interest: photonic, applied machine learning on raspberry pi system • Python user 2014-now
Python at a glance • Multipurpose • OOP (everything is
object) • Dynamic typing • Batteries included • Case sensitive
None
Things we can build computer vision search engine security tools
embedded system web services web application machine learning apps
Your First Python Code print “hello, world!” print(“hello, world!”)
Data Structure • Integer • Float • List • String
• Boolean • Dictionary • Tupple
Data Structure: list • Create an empty list • Add
an item into it • Access item • Remove some item fr = [] fr.append(“guava”) fr[0] fr.remove(“guava”)
Data Structure: tupple • Immutable array • Data are read-only
• Items in tupple can’t be deleted • Can’t add more data once it’s created fixed_data = (1, 2) fixed_data[0] # 1 fixed_data[0] = 1
Data Structure: dict • Named-list • Key-value user = {“name”
“vicky”, “age”: 21} user[“name”] user.keys() user.values() user = {name=“vicky”, age=21}
Control Flow In Python, there are: • if • elif
• else • for • while • continue • break
Control Flow: if if 1 > 0: print “1 larger
than 0”
Control Flow: if-else if 1 > 2: print “hola” else:
print “holi”
Control Flow: if-elif-else if 1 > 2: print “hola” elif
1 > 3: print “holu” else: print “holi”
Loop: for • for loop is for iterating over iterable
object • range function creates list which is iterable • in above case, the i is 0, 1, 2, ..., 99 on each iteration for i in range(100): print(i)
Looping: while • while Requires a condition in order to
start or terminate the loop (while-break) while True: print “hello!” N = 0 while N < 10: N += 1
Function: def • def is the keyword for creating function
def gravity_force(M, m, r): return 6.62e+12*(M*m)/r**2 Function keyword Returned value(s) arguments function name
Coding style in Python • Python uses indentation • 3
spaces or 1 tab for each scope • No brackets or semicolons!
Python Application Case: Machine Learning Security
Machine Learning: Security: Predict land price by its size Port
scanner
Sec: Port scanner • Door knocking analogy • Bruteforce
ML: Linear Regression • Linear model out = mx+b •
Scikit learn search for the m • so the predicted value can be close to y
ML: Datasets X (M2) Y ($ USD) 200 2000 250
3500 300 5000 378 5400 456 6500 680 8700 800 10000
ML: model training datasets model f(x)
ML: prediction new input predicted output Trained Model
ML: scikit-learn • Machine learning is difficult problem • Fortunately,
we have scikit-learn
Python resources • github.com/vinta/awesome-python