10
Sophomore Slumpware Sophomore Slumpware Predicting Album Sales Predicting Album Sales with Artificial Neural with Artificial Neural Networks Networks Matthew Wirtala ECE 539 Matthew Wirtala ECE 539

Sophomore Slumpware

  • Upload
    kevork

  • View
    24

  • Download
    4

Embed Size (px)

DESCRIPTION

Sophomore Slumpware. Predicting Album Sales with Artificial Neural Networks Matthew Wirtala ECE 539. Overview. Record sales have decreased ~30% over the past 4 years No consensus on why this is File-sharing? Inferior albums being released?. Overview. - PowerPoint PPT Presentation

Citation preview

Page 1: Sophomore Slumpware

Sophomore SlumpwareSophomore Slumpware

Predicting Album Sales with Predicting Album Sales with Artificial Neural NetworksArtificial Neural Networks

Matthew Wirtala ECE 539Matthew Wirtala ECE 539

Page 2: Sophomore Slumpware

OverviewOverview

Record sales have decreased ~30% over Record sales have decreased ~30% over the past 4 yearsthe past 4 years

No consensus on why this isNo consensus on why this isFile-sharing?File-sharing? Inferior albums being released?Inferior albums being released?

Page 3: Sophomore Slumpware

OverviewOverview

Perhaps album sales can be predicted Perhaps album sales can be predicted with an MLP networkwith an MLP networkMay show what factors determine how well an May show what factors determine how well an

album will sellalbum will sell Indicate which albums deserve a better Indicate which albums deserve a better

marketing pushmarketing push

Page 4: Sophomore Slumpware

Feature dataFeature data

Critical acclaimCritical acclaimReview scores gathered from 4 sourcesReview scores gathered from 4 sources

www.pitchforkmedia.comwww.pitchforkmedia.comwww.allmusic.comwww.allmusic.comwww.metacritic.comwww.metacritic.comRolling StoneRolling Stone

Page 5: Sophomore Slumpware

Feature dataFeature data

Hype levelHype levelAmount of press coverage will lead to higher Amount of press coverage will lead to higher

public awareness and possibly higher album public awareness and possibly higher album salessales

Previous album salesPrevious album salesServe as barometer of how established an Serve as barometer of how established an

artist may be. artist may be.

Page 6: Sophomore Slumpware

Data labellingData labelling

Too difficult to predict exact album salesToo difficult to predict exact album sales

Data labelled as one of three classesData labelled as one of three classesAlbums that sell fewer than 500,000 copiesAlbums that sell fewer than 500,000 copiesGold albums (500,000 – 1,000,000 copies)Gold albums (500,000 – 1,000,000 copies)Platinum albums ( > 1,000,000 copies sold)Platinum albums ( > 1,000,000 copies sold)

Page 7: Sophomore Slumpware

Data preprocessingData preprocessing

Data gathered for 60 albumsData gathered for 60 albums20 from each class20 from each classSome from same artist falling into separate Some from same artist falling into separate

classesclassesData randomized and split into three Data randomized and split into three

partitionspartitionsFeature vectors normalized to -5 - +5Feature vectors normalized to -5 - +5

Page 8: Sophomore Slumpware

The Neural NetworkThe Neural Network

Utilized Professor Hu’s standard bp.m Utilized Professor Hu’s standard bp.m algorithmalgorithm

Trialed many different configurationsTrialed many different configurationsOptimal configurationOptimal configuration

2 hidden layers2 hidden layers7 neurons in first layer, 8 in second7 neurons in first layer, 8 in secondLearning rate = 0.267, momentum = 0.007Learning rate = 0.267, momentum = 0.007Tested with 3-way cross validationTested with 3-way cross validation

Page 9: Sophomore Slumpware

ResultsResults

Highest classification rate 60%Highest classification rate 60%Correctly classified class 1 and 2 albums with Correctly classified class 1 and 2 albums with

80-90% accuracy80-90% accuracyCould not separate class 2 albumsCould not separate class 2 albums

Class 2 featured albums with vectors similar to Class 2 featured albums with vectors similar to those of classes 1 and 3those of classes 1 and 3

Sample confusion matrix:Sample confusion matrix: 4 0 24 0 2 2 0 52 0 5 1 0 61 0 6

Page 10: Sophomore Slumpware

Future ImprovementsFuture Improvements

Further analysis of feature vectors to Further analysis of feature vectors to determine possible differences in class 2 determine possible differences in class 2 albumsalbums

Possible reduction of labelling to two Possible reduction of labelling to two classes (combine Gold and Platinum)classes (combine Gold and Platinum)

Classification does show that predictions Classification does show that predictions can be made based on the features can be made based on the features considered in this studyconsidered in this study