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
Introduction to Machine Learning
Search
Tiago Martinho
May 01, 2018
Technology
0
46
Introduction to Machine Learning
Tiago Martinho
May 01, 2018
Tweet
Share
More Decks by Tiago Martinho
See All by Tiago Martinho
Time Managment
tiagomartinho
0
43
BuddyBuild
tiagomartinho
0
38
Daily Journal
tiagomartinho
0
54
Everyone can code
tiagomartinho
0
36
Silicon Valley Tour
tiagomartinho
1
69
Automated User Interface Testing
tiagomartinho
0
64
Swift Peer Lab - try! Swift Tokyo
tiagomartinho
0
89
Francigenr
tiagomartinho
1
35
Artusi Learning
tiagomartinho
0
44
Other Decks in Technology
See All in Technology
Contract One Engineering Unit 紹介資料
sansan33
PRO
0
7k
新規事業におけるAIリサーチの活用例
ranxxx
0
130
低レイヤソフトウェア技術者が YouTuberとして食っていこうとした話
sat
PRO
7
5.8k
AI Ready API ─ AI時代に求められるAPI設計とは?/ AI-Ready API - Designing MCP and APIs in the AI Era
yokawasa
20
5.7k
DATA+AI SummitとSnowflake Summit: ユーザから見た共通点と相違点 / DATA+AI Summit and Snowflake Summit
nttcom
0
190
OTel 公式ドキュメント翻訳 PJ から始めるコミュニティ活動/Community activities starting with the OTel official document translation project
msksgm
0
200
Ktor + Google Cloud Tasks/PubSub におけるOTel Messaging計装の実践
sansantech
PRO
1
230
Microsoft Defender XDRで疲弊しないためのインシデント対応
sophiakunii
3
400
[SRE NEXT 2025] すみずみまで暖かく照らすあなたの太陽でありたい
carnappopper
2
880
20250718_ITSurf_“Bet AI”を支える文化とコストマネジメント
helosshi
1
210
AWS Well-Architected から考えるオブザーバビリティの勘所 / Considering the Essentials of Observability from AWS Well-Architected
sms_tech
1
850
分散トレーシングによる コネクティッドカーのデータ処理見える化の試み
thatsdone
0
170
Featured
See All Featured
How GitHub (no longer) Works
holman
314
140k
Helping Users Find Their Own Way: Creating Modern Search Experiences
danielanewman
29
2.8k
Code Reviewing Like a Champion
maltzj
524
40k
Fantastic passwords and where to find them - at NoRuKo
philnash
51
3.3k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
47
9.6k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
367
26k
For a Future-Friendly Web
brad_frost
179
9.8k
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
kevinliebholz
34
5.9k
A Tale of Four Properties
chriscoyier
160
23k
Automating Front-end Workflow
addyosmani
1370
200k
The Cult of Friendly URLs
andyhume
79
6.5k
Put a Button on it: Removing Barriers to Going Fast.
kastner
60
3.9k
Transcript
Tiago Martinho @martinho_t tiagomartinho Introduction to Machine Learning
What is ML?
Computer science Artificial Intelligence Machine Learning Pattern Recognition and Computational
Learning Theory
"the ability to learn without being explicitly programmed” Arthur Samuel,1959
"A computer program is said to learn from experience E
with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.” Tom M. Mitchell
Task Detecting Handwriting Characters
Task Experience Detecting Handwriting Characters Labelled Handwriting Characters
Task Performance Experience Detecting Handwriting Characters Detects Characters w/ Higher
Accuracy Labelled Handwriting Characters
Why ML?
MNIST simple computer vision dataset ML Hello World
None
28x28 = 784 numbers
uses the examples to automatically infer rules for recognising handwritten
digits 0 1 2 3 4 5 6 7 8 9
ML Applications
Fraud Detection Self-Driving Cars OCR Search engines Computer Vision Health
Monitoring … NLP
OrCam http://www.orcam.com
Alpha Go https://techcrunch.com/2017/05/23/googles-alphago-ai-beats-the-worlds-best-human-go-player/
Poker https://www.scientificamerican.com/article/time-to-fold-humans-poker-playing-ai-beats-pros-at-texas-hold-rsquo-em/
How it works
Supervised Learning
Supervised Learning
Supervised Learning General Rule Y = M*x + b
Supervised Learning
Unsupervised Learning
Unsupervised Learning
Unsupervised Learning
Support Vector Machine
SVM
Anomaly detection
Anomaly detection
Anomaly detection
Anomaly detection
Training Inference
Features
Collect Train Classify
Data 1. Train (60%) 2. Test (20%) 3. Validation (20%)
Can we generalise?
None
None
None
Tiago Martinho @martinho_t tiagomartinho Thank you!