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
Exercises for Patterns in Recordings
Search
Ben Fields
July 05, 2016
Technology
0
51
Exercises for Patterns in Recordings
from DHOxSS 2015-16
Ben Fields
July 05, 2016
Tweet
Share
More Decks by Ben Fields
See All by Ben Fields
People in the loop machine learning: A case Study in news similarity
bfields
0
160
Human-centric evaluation of similarity spaces of news articles
bfields
2
79
Ethics, Data Science, and Public Service Media
bfields
1
78
The Case for Public Service Recommender Algorithms
bfields
0
780
Bikes are Dope
bfields
0
85
Beyond Your Reckons: from feels to facts
bfields
0
91
People who like cheese also like crackers: a learning hour on recommender systems
bfields
0
98
Auto-Summarising Beer Reviews
bfields
0
80
rMIXr: how we learned to stop worrying and love the graph
bfields
0
110
Other Decks in Technology
See All in Technology
Introduction to Bill One Development Engineer
sansan33
PRO
0
320
その意思決定、まだ続けるんですか? ~痛みを超えて未来を作る、AI時代の撤退とピボットの技術~
applism118
42
24k
リアーキテクティングのその先へ 〜品質と開発生産性の壁を越えるプラットフォーム戦略〜 / architecture-con2025
visional_engineering_and_design
0
7.7k
社内外から"使ってもらえる"データ基盤を支えるアーキテクチャの秘訣/登壇資料(飯塚 大地・高橋 一貴)
hacobu
PRO
0
7.7k
生成AIシステムとAIエージェントに関する性能や安全性の評価
shibuiwilliam
1
210
一億総業務改善を支える社内AIエージェント基盤の要諦
yukukotani
1
240
AI駆動開発2025年振り返りとTips集
knr109
1
120
クラスタ統合リアーキテクチャ全貌~1,000万ユーザーのウェルネスSaaSを再設計~
hacomono
PRO
0
190
機械学習を「社会実装」するということ 2025年冬版 / Social Implementation of Machine Learning November 2025 Version
moepy_stats
4
500
履歴テーブル、今回はこう作りました 〜 Delegated Types編 〜 / How We Built Our History Table This Time — With Delegated Types
moznion
0
380
2025 DORA Reportから読み解く!AIが映し出す、成果を出し続ける組織の共通点 #開発生産性_findy
takabow
0
420
Contract One Engineering Unit 紹介資料
sansan33
PRO
0
9.7k
Featured
See All Featured
What's in a price? How to price your products and services
michaelherold
246
12k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
231
22k
Unsuck your backbone
ammeep
671
58k
Reflections from 52 weeks, 52 projects
jeffersonlam
355
21k
GraphQLの誤解/rethinking-graphql
sonatard
73
11k
How to Think Like a Performance Engineer
csswizardry
28
2.3k
The World Runs on Bad Software
bkeepers
PRO
72
12k
Bootstrapping a Software Product
garrettdimon
PRO
307
110k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
21
1.3k
Git: the NoSQL Database
bkeepers
PRO
432
66k
The Myth of the Modular Monolith - Day 2 Keynote - Rails World 2024
eileencodes
26
3.2k
Into the Great Unknown - MozCon
thekraken
40
2.2k
Transcript
Patterns in Recording - Exercises Ben Fields
WEKA basics • Launch the WEKA Application • Select ‘Explorer’
• Load ’50_weka_class_labeled.arff’ • Select various Attributes (features)
WEKA basics • Attribute selection with regular expressions (‘.*’ is
an expanding wild card) • Find all the MFCC attributes • also select ‘CLASS’ • press ‘Invert’, then ‘Remove’
using a classifier • Select ‘Classify’ tab • ‘Choose’ >>
‘classifiers/bayes/ NaiveBayes’ • set test options to ‘Cross-validation’ • press ‘Start’
putting it all together • now repeat the whole process
using LPC features
Explore more files and features • Load ‘training_data_after_parsing.arff’ • Classify
against the CLASS mood labels using Random Forests and J48. Which performs better? Which classes are most confused?